|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
RECRUITMENT OF TRANSCRIPTION COMPLEXES TO THE BETA-GLOBIN
LOCUS IN VIVO AND IN VITRO
KAREN FRANCES VIEIRA
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
Karen Frances Vieira
This document is dedicated to my mother and sister who have been there for me every
step of the way, and to my father who is gone but not forgotten.
I would like to first acknowledge God as the guiding light in my life, for without
Him my world would be empty.
I owe a huge thanks to my mentor, Dr. Jorg Bungert, for the opportunity to work in
his lab and for receiving the excellent training that I did. He has been a great mentor,
motivator, advisor and teacher. He has been patient with me and very supportive of my
alternate career choices. He is the hallmark of a good scientist, and I respect and am
inspired by his commitment and dedication.
I thank all of my committee members, Drs. Michael Kilberg, Thomas Yang, Linda
Bloom, Hideko Kasahara, and James Resnick. They have all been a great help in
providing suggestions and directions for my experiments and career. Drs. Kilberg and
Yang have been especially supportive and encouraging throughout my time at the
University of Florida, and have always had insightful comments during my committee
meetings and presentations. I need to thank them as well as Dr. Brian Cain for critically
evaluating and helping me with my presentation that won first place in the 2004 Medical
Guild Research Competition. I appreciate Dr. Resnickjoining my committee at such
short notice, and also providing useful information to help me get on track for graduation.
I extend my appreciation to Dr. James B. Flanegan, the department chair, and all
the administrative and secretarial staff that has made me at home in the department. I
would like to say a special thanks to Bradley Moore who always shares a smile or a
compliment to brighten my day.
I thank all of the past and present members of the Bungert lab. When I first started
in the laboratory, Kelly Leach and Padraic Levings instantly made me feel comfortable
and were great mentors to me. I appreciate Kelly for setting the standards, which I have
since tried to meet or surpass. Sung-Hae Lee Kang also was a mentor and a great source
of encouragement and advice as well as a great friend. Ara Aslanian is the post-doctoral
researcher who, upon leaving, passed on his protocols and projects to me. Since that was
the first step in my successes here at the University of Florida, I owe Ara many thanks.
Christof Dame was a post-doc in the laboratory that encouraged me with helpful,
thoughtful and sometimes amusing discussions. I would like to thank Valerie Crusselle
for being a great colleague and also a friend. I wish her many blessings. Takeesha
Roland has been the best laboratory technician I know and I appreciate her help, patience
and sincere friendship. I need to thank the many other members of the lab including
Boris Thurisch, Felicie Andersen, Meredith Hill, and the many undergraduates and high
school students who have helped and entertained me along the way.
I thank the members of Dr. Yang's lab who have always been helpful and friendly,
especially Christine Kiefer and Sara Rodriguez, who both helped me with footprinting
experiments. I would like to say a special thanks to the members of Dr. Kilberg's lab,
especially Hong Chen who has been a great source of knowledge.
I would like to thank my family and friends who have been very supportive. My
mother and sister have always been my best source of support and encouragement, and I
want them to know that I love and appreciate them. I thank my boyfriend and many
friends, who have helped me tremendously through the stresses of graduate school, and
have been patient with my many moods during this long journey.
TABLE OF CONTENTS
A C K N O W L E D G M E N T S ................................................................................................. iv
LIST OF FIGURES ......... ......................... ...... ........ ............ ix
ABSTRACT .............. .......................................... xi
1 IN TR OD U CTION ............................................... .. ......................... ..
H em oglobin and Sickle Cell D disease ........................................ ....................... 1
Overview of Sickle Cell and Thalassemia.................................. ...............
Sym ptom s and Treatm ents ............................................................................2
Current R research on Therapies ........................................ ......... ............... 3
C hrom atin Structure........ ....... .............................................................. ....... .. .. ....
H stone M modifications ....................... ............................ .. ........ ............ .....
C hrom atin R em odeling ............................................................................. 7
M echanism of Transcription............................................................ ............... 8
Preinitiation Com plex A ssem bly.................................................. .................. 9
In itia tio n ................................................................................................................ 9
The P-G lobin G ene Locus ............................ ............... ..................... 1
L location and O organization ............. .............................................. ... ............. 11
Locus Control Region..................................................... 12
Proteins Associated with the P-Globin Locus ......... .......... .....................14
Model for 3-Globin Gene Regulation......... ................. ...........................16
Questions to Be Addressed........... .......... ........... .................. ............... 19
2 M ATERIALS AND M ETHOD S ........................................ ......................... 21
Cell Culture........................................ ......... 21
Chrom atin Im m unoprecipitation (ChIP) ......... ..... ............ ................ ..................21
Chromatin Immunoprecipitation and DMS Footprinting ..........................................24
D M S Treatm ent ....... ........ ....................... .......... ..... ........ ........ .... 26
Linker LM PCR ....... ............................ .......... ... .. ................. 26
W western Blotting ........................ ....... ......... ................ ....... .. 29
C ell Synchronization .............................................................30
PCR-Based DNA Replication Analysis................................... ....................... 30
D N A Isolation ............................................ .. .. ............. ......... 30
Sem i-Q uantitative PCR ......................................................... ..................... 31
In Vitro Polym erase Transfer Analysis.................................... ....................... 32
MEL Cell DMSO Differentiation..................... ....... ............................ 33
Sem i-Q uantitative R T-PCR ............................................. .............................. 33
D double Im m unoprecipitation........................................................... ............... 34
Co-Im m unoprecipitation..............................................................................36
3 CHROMATIN IMMUNOPRECIPITATION AND DMS FOOTPRINTING ...........38
In tro d u ctio n ............................ ....................................... ................ 3 8
R e su lts ............................................................................................ 4 0
D iscu ssio n ......................................................................................... 4 3
4 INTERACTION OF RNA POLYMERASE II AND TRANSCRIPTION FACTORS
W ITH THE P-GLOBIN LOCU S ........................................ ......................... 46
In tro d u ctio n .......................................................................................4 6
R e su lts ...........................................................................................4 8
D isc u ssio n ............................................................................................................. 5 8
5 ANALYSIS OF THE MECHANISM OF RNA POLYMERASE II TRANSFER
FROM THE LCR TO THE 3-GLOBIN GENE PROMOTER ................65
In tro d u ctio n ........................................................................................................... 6 5
R e su lts ...........................................................................................6 6
D isc u ssio n ............................................................................................................. 7 2
6 DETERMINING THE FUNCTION OF THE USF, TFII-I AND HDAC3
PROTEINS IN 3-GLOBIN GENE REGULATION ................................................76
Introdu action ............... ................ ................................................................... 76
Results ................... ..................................78
D discussion ............... ...... ...................................................... ........82
7 CONCLUSIONS AND FUTURE DIRECTIONS ...................................... 87
C h IP F o o tp rin tin g ......................... ....................................................................... 8 7
ChIP w within the 3-G lobin Locus................................................. 88
C ell Synchronization ............................................................ 89
Chromosome Conformation Capture ....... ...................... .......... ............ 90
In Vitro Pol II Transfer ....................... .... .................. 91
USF, TFII-I and HDAC3 Function in 3-Globin Regulation .....................................92
Summary ................... .................................. 93
L IST O F R E F E R E N C E S ...................................... .................................... ....................94
BIOGRAPHICAL SKETCH ............................................................. ..................113
LIST OF FIGURES
1-1. The hum an and m house P-globin loci ..................................................................... 13
1-2. Multistep model for human 3-globin gene regulation.............................................18
3-1. Sequence alignment of the adult P-globin downstream promoter region. ...............41
3-2. Analysis of protein-DNA interactions in the murine P-globin downstream
promoter region by a combination of chromatin immunoprecipitation and DMS
footprinting.......................................... .................... ...... ......... 42
3-3. ChIP experiment showing the interaction of USF and NF-E2 with the murine
pmaj-globin promoter in M EL cells............................ ...... ............. .... 43
3-4. LMPCR footprint analysis of non-selected or ChIP-selected chromatin ...............44
4-1. Diagrammatic representation of the human and murine P-globin gene locus. ........49
4-2. Interaction of transcription factors and RNA polymerase II with the human 3-
globin locus and the human necdin gene. ..................................... ............... 50
4-3. Interaction of transcription factors with and transcription in the murine P-globin
locu s in M E L cells............ ... ........................................................ .......... .. .... 52
4-4. Interactions of transcription factors and RNA polymerase II with the human 3-
globin locus during early S phase in synchronized K562 cells..............................54
4-5. Interactions of transcription factors and RNA polymerase II with the human 3-
globin locus during early S phase in synchronized MEL cells. ............................56
4-6. Semi-quantitative PCR analysis of DNA content of synchronized K562 cells. ......57
4-7. Semi-quantitative PCR analysis of DNA content of synchronized MEL cells........58
5-1. A schematic representation of the in vitro Pol II transfer experimental procedure. 67
5-2. Transfer of RNA polymerase II from immobilized LCR constructs to the
P -g lo b in g en e ............ ...... ... .... ................. ............. ................ 6 8
5-3. Analysis of the specificity of Pol II transfer. ............. .............................................70
5-4. Comparison of Pol II transfer using the wild-type P-globin gene (lane 1) and a
mutant P-globin gene (lane 2) in which an E-box at +60 was altered....................70
5-5. Transfer of RNA polymerase II to the P-globin gene in the presence of NF-E2 or
B SA ............................................................................... 7 1
5-6. Analysis ofNF-E2 mediated transfer to the P-globin gene...................................72
5-7. Quantitative summary of the transfer experiments. .............................................. 73
5-8. Model of transcription complex recruitment to the P-globin gene .......................75
6-1. Sequence alignment of the human 3-globin downstream promoter region. ...........78
6-2. Characterization of protein-DNA interactions in the human 3-globin
downstream promoter region in vivo........ ..................................... ............... 80
6-3. Chromatin double immunoprecipitation (ChDIP) experiment analyzing the
interaction of TFII-I and HDAC3 at different regions within the human 3-globin
lo cu s. .............................................................................. 8 1
6-4. Co-immunoprecipitation experiment analyzing the interaction of TFII-I and
H D A C3 as a com plex ............................................... .. ...... .. ............ 82
6-5. Summary and model of protein-DNA interactions at the human 3-globin
prom other. ............................................................................86
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
RECRUITMENT OF TRANSCRIPTION COMPLEXES TO THE BETA-GLOBIN
LOCUS IN VIVO AND IN VITRO
Karen Frances Vieira
Chair: Jorg Bungert
Major Department: Biochemistry and Molecular Biology
The P-globin gene locus is the site of many mutations and deletions that cause
sickle cell disease and P-thalassemias. Erythroid-specific, high-level expression of the 3-
globin genes is regulated by the locus control region, located far upstream of the genes.
Recent studies show that locus control region core elements recruit RNA polymerase II.
Here, we developed and optimized methods to study protein-DNA interactions
within the P-globin locus. These techniques were used to analyze the interaction of
transcription factors and RNA polymerase II with the P-globin locus in erythroleukemia
cell lines. The data show that transcription initiation complexes are recruited to the LCR
and to the genes. Moreover, RNA polymerase II dissociates during early S phase in
synchronized erythroid cells suggesting that replication disrupts the association of
transcription complexes with the globin locus. The interaction of NF-E2 and USF2
precedes the re-association of RNA polymerase II implicating these proteins in a process
regulating the recruitment of RNA polymerase to the globin locus during replication.
Furthermore, we demonstrate transfer of transcription complexes from immobilized
LCR constructs to the human 3-globin gene in vitro. Our data are consistent with the
hypothesis that the LCR and the genes cooperate to recruit transcription complexes to the
globin gene locus.
Studies on the interaction of proteins with the adult P-globin gene promoter
demonstrate that the helix-loop-helix proteins USF and TFII-I associate with DNA-
regulatory elements located at or downstream of the transcription initiation site. TFII-I
interacts with histone deacetylase 3 and binds more efficiently in erythroid cells that do
not express the P-globin gene. In contrast, USF, known to interact with chromatin-
opening activities, associates with the globin gene when it is expressed. These data
suggest that USF and TFII-I regulate P-globin gene expression in an antagonistic fashion.
Hemoglobin and Sickle Cell Disease
Hemoglobin is the molecule in red blood cells that transports oxygen to all the
tissues of mammalian organisms. Hemoglobin is composed of two alpha (a) globin
chains and two beta (3) globin chains, each carrying a heme molecule . The heme
molecule contains an iron atom and is responsible for binding the oxygen that is
transported. The a- and P-globin chains are important for the overall structure of
hemoglobin, and thus contribute to its essential functions in gas transport.
Overview of Sickle Cell and Thalassemia
More than 80 point mutations and deletions in the a-globin genes are known to
cause a-thalassemia . These thalassemias have the highest incidence in India, Africa,
and Arabic countries . Similarly, more than 200 mutations and deletions in the 3-
globin genes cause sickle cell anemia, sickle SC disease, sickle SD disease (collectively
called sickle cell disease), and P-thalassemias . Sickle cell disease and the 3-
thalassemias are the most common monogenetic diseases worldwide, with an estimated
80 million P-thalassemia carriers . There are over 10,000 African-Americans with
sickle cell anemia in the USA alone. In African, Asian, Indian and Middle Eastern
communities, sickle cell disease and P-thalassemia are widespread.
Because some mutations associated with the a- and P-globin gene loci provide
protection from malaria, incidence for thalassemias and sickle-cell disease are historically
more common in areas affected by malaria. These regions include Africa, India, and
Mediterranean countries. Heterozygous individuals carrying the sickle cell allele have a
survival advantage since these individuals are generally healthy but reveal resistance to
malaria. In the past individuals homozygous for the disease generally did not live past
childhood. These diseased individuals now benefit from an extended lifespan due to the
increased level of worldwide healthcare.
Symptoms and Treatments
Sickle hemoglobin polymerizes when deoxygenated. The hemoglobin polymers
form alpha helical bundles, which deform red blood cells into a sickle-shape. These
sickle cells that give the disease its name are more rigid than normal red blood cells and
cannot pass through small blood vessels. The resulting blockage of blood vessels causes
intense pain and other side effects.
Patients with sickle cell disease experience a range of complications, including
painful crises, acute chest syndrome, priapism in males, stroke, bone damage, and lung
damage. Many patients undergo monthly blood transfusions to lower the percentage of
sickle red blood cells in their system. This reduces the frequency of painful crises, but
also increases iron concentrations to sometimes-toxic levels. Therefore, along with
transfusions, sickle cell patients undergo iron chelation therapy. Unfortunately, since
chelation therapy is performed by daily intravenous injections there is a low patient
compliance. Though, the recent development of new oral iron-chelating agents may
allow for better compliance .
Currently, the only cure for severe hemaglobinopathies is hematopoietic stem cell
transplantation (HSCT). About 71% ofthalassemia patients given HSCT from a matched
donor have been cured of their disease . Due to the high risks associated with this
procedure, patients receive HSCT only if sickle-related death is imminent .
Encouragingly, of the sickle cell patients that have received HSCT, 85% survived
disease-free . Since only about 1% of sickle cell patients meet the criteria for HSCT
and have a suitable donor, alternative therapies are highly sought after.
Current Research on Therapies
Many current trials are performed with drugs that increase the levels of fetal
hemoglobin (HbF) since HbF inhibits the polymerization of sickle hemoglobin (HbS).
The drug 5-azacytidine, which inhibits DNA methyltransferases, has been used in the
past but due to serious side effects is now rarely used. Sodium butyrate acts
synergistically with 5-azacytidine and reactivates y-globin expression for the production
of HbF . Hydroxyurea, which is known to inhibit DNA synthesis and alter chromatin
structure, is currently the drug of choice to raise HbF in sickle cell patients and to reduce
symptoms associated with the disease [9, 10]. The main known side effect of
hydroxyurea treatment is myelotoxicity, or destruction of the bone marrow. Its long-term
toxicity and effects on survival are yet to be determined. For this and other reasons,
hydroxyurea is reserved only for patients with severe complications.
For patients that do not meet the criteria for HSCT or do not respond to
pharmacological therapies, gene therapy may provide a promising alternative in the
future. Sickle cell disease is caused by a point mutation in the coding region of the 3-
globin gene and it was the first genetic disorder in which such point mutations have been
found . Due to the relative simple genetic defect and the ability to isolate and
transfect hematopoietic stem cells, it was anticipated that globin gene disorders would be
among the first for which gene therapy approaches are feasible.
Several viruses are being tested as vectors for therapeutic gene transfer of the 3-
globin genes to mammalian cells. These include murine retroviruses , lentiviruses
, human foamy virus , and adeno-associated virus [15-18]. Promising results
were obtained using lentivirus-based gene therapy in a mouse model for P-thalassemia
, though there is still much to be done before it can become approved as a safe form
of gene therapy for sickle cell and P-thalassemia.
One of the obstacles of using lentivirus is the fact that the DNA integrates into the
genome and there is some evidence that integration into open transcriptionally active
regions is preferred [20, 21]. This enhances the risk of gene disruption (e.g., of tumor
suppressor genes) or activation of endogenous genes (e.g., oncogenes). Adeno-associated
virus is considered to be the safest virus used in gene therapy trials because it does not
appear to cause strong immune reactions and in many reports the DNA remained
episomal [22-25]. The limitation of using AAV for globin gene therapy is the low
packaging capacity. As outlined below, appropriate control of P-globin gene regulation
is complex and requires the presence of many cis-acting DNA elements. Alternative
approaches are to optimize HSCT for making it more broadly applicable and to use novel
mechanisms to induce a high level of fetal globin gene expression.
To comprehend the regulation of the globin genes and potential interactions within
the P-globin gene locus, it is important to understand the chromatin structure in the locus
and its relevance to gene expression.
In order to fit the entire genome into the nucleus of each living cell DNA is highly
constrained and compacted through interactions with proteins called histones. Histones
are small basic proteins consisting of a globular domain and a more flexible N-terminus.
DNA organization in the nuclei of eukaryotes involves repetitive packaging into
nucleosomes, which are composed of two each of the histone core proteins H2A, H2B,
H3 and H4 arranged in an octamer around which 146 base pairs of DNA is wrapped [26,
27]. Nucleosomal DNA and its associated proteins interact with other proteins to form
higher order structures in the nucleus; the combination of DNA and associated factors in
the nucleus is termed chromatin.
Nucleosomal arrays along the DNA are proposed to fold into a 30 nm fiber upon
incorporation of the linker histone H1, though the mechanism of this compaction remains
unclear. Higher levels of compaction may occur in genomic domains termed
heterochromatin. Genes located in these dense chromatin domains are generally not
accessible for interacting with transcription factors and are thus suppressed [28, 29].
Even further compaction is required during mitosis, and a protein called condensin helps
convert chromatin into condensed mitotic chromosomes . Conversely, there are
regions of the genome, termed euchromatin, in which the genes are not highly
compacted, leaving them accessible to transcription complexes required for gene
Although the details of chromatin folding are still unclear, the relative compaction
of any region of the genome can be measured by determining the susceptibility of that
region to digestion by the endonuclease deoxyribonuclease (DNaseI). Regions of the
genome are described as being DNaseI insensitive, sensitive, or hypersensitive.
Insensitive regions like heterochromatin are highly condensed and inaccessible to
DNaseI. Sensitive regions such as euchromatin exhibit a less compacted chromatin
structure and are accessible to DNaseI . Transcribed regions of the genome are
usually DNaseI sensitive and may be organized as a 30 nm chromatin fiber. DNaseI
hypersensitive sites are usually 200 to 400 bps long and probably reflect regions in which
nucleosomes have been removed or modified. These sites contain clusters of transcription
factor binding sites and are generally associated with the function of regulatory DNA
regions, e.g., enhancers or promoters .
Nucleosomes and other forms of DNA compaction modulate transcription and thus
represent an important factor in gene regulation . Enzymes that modify and remodel
chromatin play an important role in activation and repression of gene expression.
Chromatin modifications are post-translational modifications of the histones.
Alternatively, chromatin remodeling includes changing the location of nucleosomes,
altering histone-DNA interactions, removal of histones, and histone exchange [34-39].
Histones can be acetylated, methylated, phosphorylated, ubiquitylated, sumoylated,
and ADP-ribosylated . Originally it was thought that modifications occurred only on
the N-terminal tails of histones that protrude from the nucleosome core, but recent work
has identified modifications within the core regions as well as on the C-terminal tails
Lysine acetylation by histone acetyltransferases (HATs) has so far been the most
studied of all histone modifications [42-44]. There are 5 families of HATs that are all
large multiprotein complexes. Histone acetylation is a reversible process, catalyzed by
histone deacetylases (HDACs) , of which there are 3 distinct families. Several lysine
residues are known to be acetylated, including lysines 9, 14, 18 and 23 on H3, and lysines
5, 8, 12 and 16 on H4. Acetylated histones are usually localized to regions of actively
transcribed DNA, therefore making them a mark of active or open chromatin domains,
although there are instances where this is not true [46, 47]. Histone acetylation can also
be associated with transcriptional repression and silencing, as well as recombination [48-
Certain lysine residues can also be methylated by histone methyltransferases
(HMTs). Similarly to HATs and HDACs, there are several families of HMTs, which can
mono-, di-, or tri-methylate lysine residues. Recently, a histone demethylase called
PAD4 has been identified , disproving the theory that methylation may be
irreversible . Specific histone methylation patterns correlate with gene activation, as
is the case with methylated lysine 4 on H3 (H3K4) . But histone methylation can
also result in gene repression and heterochromatin formation, as with lysine 9
methylation on H3 (H3K9) [55, 56].
There are several other histone modifications, as mentioned at the beginning of this
section, the function of which has not all been determined. A recent model, called the
histone code hypothesis, postulates that the combinations of histone modifications
interact with other proteins that use this code to execute specific gene expression patterns
. This model explains how the same modification on different residues can have
different outcomes. It also explains how the same modification on the same residue but
in the context of different surrounding modifications can have different consequences,
such as H3 with acetylated lysines correlating with both activation and repression of gene
The energy from ATP hydrolysis can be used to loosen DNA-histone contacts in
order to mobilize histones. For this reason, all chromatin remodeling complexes contain
an ATPase subunit, which is critical for nucleosome mobilization. Thus far there are 3
known families of ATP-utilizing chromatin remodeling complexes, and possibly many
more to be discovered .
Though the mechanism of nucleosome movement has not been established, there
have been several models proposing that the mechanism involves DNA dissociating from
the nucleosome and being replaced by a neighboring piece of DNA. This would yield a
DNA loop that moves in a wave across the surface of the nucleosome, ensuring that as
the histone moves only a small region of contact with the DNA is broken at any one time.
This could also apply to histone contacts being disrupted with one piece of DNA and
reestablished simultaneously on another piece of DNA as histones are relocated within
the genome .
Chromatin remodeling complexes also have the ability to improve the accessibility
of chromatin to proteins. This would conceivably be achieved by modifying DNA-
histone contacts. Work with the glucocorticoid receptor Gal4-VP16 and some other
transcriptional activators demonstrated that the activators bind chromatin and then recruit
the chromatin remodeling complexes [58-60]. Other studies show that chromatin needs
to be remodeled before transcription factors interact with their binding sites in regulatory
regions . Further work has provided evidence that there is specificity of activators or
repressors for certain chromatin remodeling complexes (possibly for those they can
physically interact with) [62-65].
Mechanism of Transcription
Activation of gene expression ultimately results in the stable association of RNA
polymerase II (Pol II) and other components of the transcription complex with gene
promoters. The first step of transcription for any gene involves assembly of the pre-
initiation complex (PIC). After PIC formation and recruitment of Pol II, the C-terminal
domain (CTD) of Pol II is phosphorylated leading to the progression from initiation to
elongation competent transcription complexes.
Preinitiation Complex Assembly
PIC formation often begins with the binding of the TBP subunit of TFIID to the
TATA box. The binding of TFIID to the TATA box is stimulated by TFIIA, and the
TFIID/TFIIA/DNA complex is recognized by TFIIB, which serves as a bridge between
the pre-initiation complex and Pol II. Pol II is recruited to the PIC together with its
associated factor TFIIF followed by the interaction of TFIIE and TFIIH [66-68].
Evidence suggests that some activators act by recruiting TFIID and TFIIA to the
promoter, and these in turn recruit the Pol II holoenzyme containing RNA Pol II and
other general transcription factors (GTFs) that may already be pre-assembled [69, 70].
Because genes differ in their promoter structure, the formation of the initiation complex
likely varies between genes. For example TATA-less genes use additional proteins to
recruit Pol II and general transcription factors. There may be a stepwise recruitment of
factors to some promoters and recruitment of a Pol II holoenzyme to other promoters.
Phosphorylation of the CTD of Pol II signals the transition from preinitiation to
initiation of transcription. CTD phosphorylation is carried out by TFIIH (at serine 5) and
pTEFb (at serine 2), and is thought to be regulated by the mediator complex in yeast or
the CRSP complex in higher eukaryotes [71, 72]. It has been postulated that the
phosphorylation of the CTD breaks contacts between Pol II and some of the GTFs,
therefore allowing Pol II to initiate transcription . Phosphorylation at the two
different series play different functional roles, since the serine-5 and serine-2
phosphorylated Pol II complexes are observed in the early initiation and elongation stages
respectively [73, 74]. It is thought that the phosphorylation status of the CTD allows
association with proteins involved in transcription elongation, histone modification, and
For some genes chromatin remodeling plays a significant role in transcription,
either for PIC formation or after PIC formation for transcription initiation and elongation
[59, 74-77]. Chromatin alterations involved in transcriptional activation are often limited
to acetylation of histone tails by HATs, though other modifications may also be involved
but have not yet been identified.
Efficient elongation requires that the Pol II CTD remains phosphorylated during the
entire elongation process. A hyper-phosphorylated CTD is important for the interaction
of Pol II with RNA processing factors such as capping enzyme, splicing factors and
proteins for transcription termination [78, 79]. The CTD becomes de-phosphorylated
after transcription termination and is then again able to interact with the promoter . A
CTD phosphatase Fcpl has been identified that dephosphorylates the CTD [80-85],
though the mechanism of dephosphorylation remains elusive.
Several other parameters have been discovered that contribute to transcription
elongation. Histone acetylation is required for efficient transcription elongation in yeast
, and seems to be lost immediately after the passage of Pol II [87, 88]. There are
several classes of elongation factors that have been identified. For example, SII
reactivates arrested Pol II by inducing the Pol II's endonucleolytic activity [89-92]. Pafl
associates with elongating Pol II to recruit histone-modifying activities .
The P-Globin Gene Locus
The human 3-globin genes are expressed in a developmental stage and tissue-
specific manner. Regulation of chromatin structure and stage-specific recruitment of
transcription complexes play imminent roles in the expression of the globin genes.
Location and Organization
The P-globin chains of hemoglobin are coded for by the P-globin gene locus
located on chromosome 11 p5.5 in humans and chromosome 7 in mice . The human
P-globin locus contains five genes arranged in the order of their developmental
expression in erythroid cells (Figure 1-1) [95, 96]. At the 5' end is the e-globin gene,
which is expressed first during the embryonic stage in erythroid cells in the yolk sac.
Next are the YG- and yA-globin genes that are activated during the fetal stage when the site
of hematopoiesis switches from the embryonic yolk sac to the fetal liver. After birth, the
site of hematopoiesis undergoes a second switch to the bone marrow, coincident with the
expression of the 6- and P-globin genes.
The 6-globin gene is expressed in adults at less than 5% of the level of P-globin
expression. This is due to a mutation in the promoter that prevents high-level expression.
Also, there are two y-globin genes, which arose by gene duplication during evolution and
their sequences only differ at amino acid 136 .
Similar to the human 3-globin locus, the mouse P-globin locus has 4 genes whose
expression is switched during development (Figure 1-1). The sy and 3hl genes are
expressed during the embryonic stage of mouse development in the yolk sac. The pmaj
and pmin genes are expressed in fetal and adult mice in the fetal liver and later in the
The globin genes are relatively small, with three coding exons and two introns
each. The exons of all P-like globin genes code for a total of 146 amino acids. These
exons correspond to the functional domains of the proteins [98, 99]. The introns are
between 117 bp to 1264 bp in size .
The entire locus is relatively sensitive to DNaseI in erythroid cells (the only cells
that express the P-globin genes), compared to its DNaseI insensitivity in non-erythroid
cells . DNaseI sensitivity around the individual genes changes based on the
developmental stage, such that expressed genes reveal an increase in DNaseI sensitivity
. Specific histone modification patterns have been detected in the P-globin gene
locus and distinguish active from inactive domains , suggesting that chromatin
remodeling activities play a role in the stage-specific transcription of the globin genes.
Locus Control Region
In both the human and mouse P-globin locus there is a prominent element located
from 6 to 22 kb upstream of the e-globin gene called the locus control region (LCR) [104,
105]. As seen in Figure 1-1, it contains 5 core regions that reveal extremely high
sensitivity to DNaseI in erythroid cells; these are called hypersensitive sites (HS1-5).
LCRs are defined by their ability to enhance expression of linked genes in a tissue-
specific, position-independent and copy number-dependent manner . The P-globin
LCR is required for high-level expression of the genes in erythroid cells. It has been
speculated that one function of the LCR is to provide an open accessible chromatin
structure to the globin genes. However, removal of the human or mouse LCR does not
affect the overall DNaseI sensitivity of the locus [107-109]. Therefore, in the
endogenous locus the LCR does not appear to be responsible for generating a DNaseI
sensitive domain. However, it is feasible that the LCR regulates chromatin structure not
through setting up DNaseI sensitivity but through histone modifications and chromatin
remodeling . This is further supported by data showing that P-globin constructs
without the LCR suffer from position-of-integration effects when integrated as transgenes
into heterochromatin, whereas P-globin constructs with the LCR do not suffer from these
position effects .
Human $-Globin Gene Locus
LCRHS5 4 3 2 1 e Gy Ay 8 p 3' HS1
Murine --Globin Gene Locus
LCR HS 5 4 3 2 1 Ey phl pmaj pmin 3' HS 1
Figure 1-1. The human and mouse P-globin loci. The 5' hypersensitive sites of the LCR
and the 3' hypersensitive sites are shown as blue boxes. The genes
expressed during the embryonic stage of development are shown as light
blue boxes. The genes expressed during the fetal stage of development are
shown as green boxes. The genes expressed during adulthood are shown as
yellow boxes. The genes expressed in both the fetal and adult stages are
shown as orange boxes. The locus as shown here is greater than 100 kb in
size and is not drawn to scale.
Stage-specific expression of the genes in the P-globin locus is dependent on their
order relative to the LCR [112, 113]. When the genes were reversed relative to the LCR
in transgenic mice the P-globin gene was expressed at the embryonic stage, whereas s-
globin expression was abolished . It has been suggested that the genes in the locus
are competitively regulated, where only one gene from the locus can be transcribed at a
time [115, 116]. This has recently been hypothesized to be due to the fact that only one
active gene can interact with the LCR at a given time .
Evidence suggests that in tissues where P-globin genes are expressed the HS cores
interact to form an LCR holocomplex [20, 118-122]. Due to the high density of protein
binding sites within the HS cores of the LCR, there are many factors that interact with
these regions. The interactions of the proteins, not only with the HS cores but also with
other proteins in other cores, may be responsible for mediating the formation of the LCR
holocomplex. The HS core flanking regions contribute to the activity of the LCR and
may stabilize interactions between core elements .
Proteins Associated with the P-Globin Locus
GATA. The zinc finger proteins GATA-1 and GATA-2 are the only two members
of the GATA family of transcription factors that are known to be expressed in erythroid
cells . GATA-1 is one of the first erythroid proteins expressed during red cell
differentiation and it is known to associate with proteins containing histone acetylase
activity . This suggests that GATA-1 may be involved in the initial opening of the
chromatin in the P-globin locus. There are many GATA-binding sequences in LCR HS
sites and in the globin gene promoters. These sites can be bound by either GATA-1 or
Maf and p45 proteins. Small maf proteins interact with p45 like proteins and bind
to MAREs (maf recognition elements) located within HS2, HS3 and HS4. Small maf
proteins contain DNA binding and protein-protein interaction domains but lack trans-
activation domains. They heterodimerize with the related proteins NF-E2 (p45), Bachl,
NRF1 or NRF2 [126-130] via the leucine zipper. The p45-like proteins are all
structurally different, and therefore may have varying functions. Studies suggest that p45
plays an important role in the function of the LCR and in transcription of the genes [127,
131, 132]. Maf/Bachl heterodimers have been shown to simultaneously bind HS2, HS3
and HS4 and bring these cores together in vitro. The interaction between Bachl
heterodimers is mediated by the BTB/Poz domain and deletion of this domain inhibited
HS site interactions in this assay [133, 134]. Bachl/maf heterodimers function as
transcriptional repressors, whereas p45/maf dimers activate transcription. Recent
evidence suggests that MAREs in the globin locus are first bound by Bachl heterodimers.
Heme mediates the dissociation of these dimers and thus allows the interaction of the
activator p45/maf to interact with regulatory elements .
EKLF. CACCC sequences, also known as KLF (kruppel like factor) binding sites
in HS2, HS3 and globin gene promoters are conserved across many species . These
sites are bound by EKLF (erythroid kruppel like factor). EKLF is required for human 3-
globin gene expression and the formation of a DNaseI hypersensitive site in the P-globin
gene promoter [137-140]. This is due to the fact that EKLF binds to the P-globin gene
promoter and recruits chromatin remodeling factors, thus causing expression of the gene
. Similarly, there is an EKLF binding site in the e-globin gene promoter, but at the
adult stage there are proteins bound at the F promoter that prevent EKLF binding and
therefore opening of the promoter for e-globin expression .
USF. An E-box, which is a CANNTG consensus sequence, located in HS2 is
known to interact with HLH (helix loop helix) proteins such as the USF (upstream
stimulatory factor) family of proteins and Tall [143, 144]. USF is a ubiquitously
expressed transcriptional activator and usually binds to DNA as a heterodimer consisting
of USF and USF2. USF can also help recruit RNA polymerase II (Pol II) transcription
complexes to initiator elements [145, 146]. Tall is an erythroid specific member of the
HLH family of proteins, which has been shown to function during the differentiation of
hematopoietic progenitor cells, and it has recently been shown to interact with TFIIH, a
member of the basal transcription complex [147, 148].
TFII-I. The transcription factor TFII-I binds to two separate elements. First is the
pyrimidine-rich initiator (Inr) element with a consensus of YYA+ NT/AYY located over
transcription start sites [149-151]. Second is the USF recognition site, the E-box. TFII-I
has a unique structure that allows for multiple protein-protein and protein-DNA
interactions . At Inr and E-box elements TFII-I interacts both physically and
functionally with USF .
Model for P-Globin Gene Regulation
Our lab has proposed a multi-step model for human P-globin gene regulation
(Figure 1-2) . The first step towards activating the globin genes involves generating
a highly accessible LCR holocomplex. Replication may or may not be required for this
region to become DNaseI sensitive. If this step is mediated by replication, then
erythroid-specific proteins may bind to the locus after replication to prevent the formation
of inaccessible chromatin domains and perhaps to also modify histone tails. It is known
that GATA-1 is one of the earliest markers of red cell differentiation  and associates
with histone acetyltransferases; therefore GATA factors may be involved in the first step
of obtaining an open chromatin structure. Other proteins such as NF-E2 may then bind
the LCR HS sites and bend or disturb the DNA structure, leading to a highly accessible
The second step is the recruitment of chromatin-remodeling and transcription
complexes to the LCR. The proteins bound at the LCR holocomplex will recruit
chromatin remodeling complexes, coactivators, and RNA polymerase II. The recruitment
of Pol II may be mediated by HLH proteins, which are known to be involved in
transcription complex formation on several TATA-less genes [145, 146]. First 'pioneer'
RNA polymerases may associate with chromatin-remodeling activities to modify the
nucleosome structure in the locus . Next, elongation incompetent polymerases may
be recruited to the LCR to be transferred to the gene promoters where they would become
phosphorylated at the C-terminal domain and be rendered elongation competent .
Step three is the establishment of chromatin domains permissive for transcription.
The globin locus has been separated into developmental stage-specific chromatin
domains based on changes in chromatin structure, DNaseI sensitivity, histone acetylation
patterns, and the presence of intergenic transcripts [102, 103, 155]. Intergenic transcripts
are non-coding transcripts found over the entire LCR and between the globin genes [102,
156]. Intergenic transcripts of the LCR initiate upstream of or within the LCR (for
example within HS2 ) and proceed towards the genes [102, 156, 157]. If the region
containing the site of initiation of adult-specific intergenic transcripts is deleted then
DNaseI sensitivity is decreased in that region and P-globin gene expression is decreased
. If the before mentioned 'pioneer' polymerases associate with chromatin-
modifying factors then those complexes can initiate intergenic transcription and modify
the chromatin structure so as to create open chromatin domains .
A. Generslmn of a hiuehl arresible LCR hobomplen
DNmecIbud hntive d 111
I IATA-1, HATP
LCFUUS 5d 3 3*
ONMuel r sirE dakna
-CATA-1, EKLF. NF-2, HtIH prmsL .i'ntn
-trutan of HiS uopmek prm ncmprss am4
sm nerifn r if1 a it IoaL hmufer~naiust 4h
B, Rnclitnient orchromllin rEun nodaiL, rctiartr. ind iranscription completes
I I rmhalm remihrie ccln
SerlartVe ctalmpir rnmuripaCO tkmps
C. Eitmihblhrnt orthrnsomr domaln permflu ite for t i sriplio
- krrk.l.I cr..eA ip
DL Transfer or maromaleculmr proelin cDmplkxns to InIdr idnl Elobii rnc promot rs
. chrielm mntik all pli
* ructivhir cneptkE
i- -\ -a
Multistep model for human P-globin gene regulation. The model proposes
four steps involved in the regulation of gene expression in the human 3-
globin locus. (A) Generation of a highly accessible LCR holocomplex. (B)
Recruitment of necessary complexes. (C) Establishment of chromatin
domains permissive for transcription. (D) Transfer of macromolecular
protein complexes to individual globin gene promoters.
The last step is the transfer of transcription complexes to the individual globin
genes. There are two feasible models of how this is accomplished. The first is the
linking model, where complexes recruited to the LCR would be transferred to the genes
via proteins bound along the DNA [101, 115]. The looping model hypothesizes that the
LCR interacts with the individual genes by looping out intervening DNA [101, 115].
Recent evidence supporting the looping model within the P-globin locus and in other
parts of the genome came from results of experiments using the chromosome
conformation capture (3C) technique [117, 120, 158-162]. Results from our lab
demonstrating that Pol II, recruited to the LCR, can be loaded onto a downstream 3-
globin promoter is compatible with the looping model .
There are possibly many factors regulating which gene promoter the LCR transfers
transcription complexes to. Due to the complexity involved in transcriptional switching,
our model does not postulate what causes the signals for switching from embryonic to
fetal and fetal to adult.
Questions to Be Addressed
The P-globin gene locus is one of the most highly studied eukaryotic gene loci.
This is in part due to the prospects of developing gene therapy strategies for sickle cell
disease and P-thalassemia. The presence of the LCR also makes the P-globin locus a
good model system for studying long-range gene regulation. Though there is a large
body of knowledge about this gene locus, there are still many unanswered questions. It is
expected that resolving these problems will contribute to successful gene therapy trials as
well as to our understanding of long-range gene regulation.
The mechanism by which the LCR activates the genes remains to be determined.
We proposed a multi-step mechanism of LCR function leading to the activation of globin
gene expression in a developmental-stage specific manner . To test this model, we
propose to determine the order of recruitment of transcription factors, Pol II and specific
histone modification marks to the locus during transcriptional activation.
The goals of this work were all related to addressing the mechanism of LCR-
mediated activation of the P-globin genes. The first goal was to develop and optimize
methods of investigating protein-DNA interactions within the P-globin locus in vivo and
in vitro. The next goal was to use these methods to compare and contrast the proteins and
complexes interacting with the LCR, active genes, and inactive genes within the locus.
The third goal was to use synchronized erythroid cells to analyze recruitment of
transcription factors and Pol II during replication. The next goal was to use an in vitro
system to investigate the mechanism of Pol II transfer from the LCR to the P-globin gene
promoter. The last goal was to elucidate the role of USF, TFII-I and HDAC3 in the
regulation of the 3-globin genes.
MATERIALS AND METHODS
All cell culture reagents were purchased from Cellgro. MEL cells were grown in
RPMI 1640 with L-glutamine supplemented with 10% fetal bovine serum and 1%
antibiotic-antimycotic (ABAM) in 5% CO2 at 370C. K562 cells were obtained from
ATCC, Coriell Repositories, and DSMZ. K562 cells were grown in RPMI 1640 with L-
glutamine supplemented with 15% fetal bovine serum and 1% penicillin-streptomycin in
5% CO2 at 370C.
Chromatin Immunoprecipitation (ChIP)
The ChIP assay was performed as described by Forsberg et al.  with minor
modifications. MEL and K562 cells (107 cells per antibody used) were grown in RPMI
medium. Crosslinking of proteins and DNAwas induced by incubating the cells in 1%
formaldehyde with rocking at room temperature for 10 minutes. The crosslinking was
quenched with 0.125 M glycine for 5 minutes. Cells were washed twice with cold lx
PBS with Complete protease inhibitors (Roche), resuspended in swelling buffer (5 mM
PIPES pH 8.0, 85 mM KC1, 0.5% NP-40, protease inhibitors) and incubated on ice for 10
minutes. The nuclei were pelleted by centrifugation for 5 minutes at 5000 rpm and 40C.
Nuclei were lysed by incubation in lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-
HC1 pH 8.1, protease inhibitors) on ice for 10 minutes. DNA was sonicatedin an ice
water bath using a Fisher model 100 sonicator at power 5 for 4 pulses of 10 seconds to
yield an average size of -500 bp. After centrifugation for 10 minutes at 14000 rpm and
4C, the supernatant was diluted in dilution buffer (0.01% SDS, 1.1% Triton X-100, 1.2
mM EDTA, 167 mM Tris-HCl pH 8.1, 167 mM NaC1, protease inhibitors) to a final
volume of x ml where x = # of antibody samples needed. 50 pl of Protein A-Sepharose
beads (Amersham) per 107 cells used were added to the diluted lysate and incubated for 2
hours with rotation at 40C. The beads were pelleted and 5 pl of appropriate antibody was
added to each ml of aliquoted supernatant and incubated overnight with rotation at 40C.
The antibodies used were TFIIB sc-225, TFIID (TBP) sc-273, Pol II (N-20) sc-899, NF-
E2 (C-19) sc-291, USF2 (C-20) sc-862 (all purchased from Santa Cruz), phospho-Pol II
05-623 and acetyl-H3 06-599 (Upstate). All antibodies were tested in Western blotting
experiments using MEL or K562 nuclear extracts as described by Leach et al .
Protein A-Sepharose beads were blocked by resuspending them in blocking buffer (3%
BSA, 0.05% Na azide in 10 mM Tris-HCl pH 8.1, 1 mM EDTA) and incubated overnight
with rotation at 40C.
The next day, the samples were rotated for 2 hours at 40C with 60 ptl blocked
Protein A-Sepharose beads. The beads were pelleted and supernatants removed. 500 ptl
of the supernatant of the 'no antibody' sample was kept and labeled 'input'. The pelleted
beads were washed with low salt wash (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 200
mM Tris-HCl pH 8.1, 150 mM NaC1), high saltwash (0.1% SDS 1% Triton X-100, 2
mM EDTA, 200 mM Tris-HCl pH 8.1, 500 mM NaC1), LiCl wash (0.25 M LiC1, 1% NP-
40, 1% Na deoxycholate, 1 mM EDTA, 100 mM Tris-HCl pH 8.1) and two washes with
TE pH 8. DNA was eluted off the beads with two volumes of 250 tl 1% SDS and 0.1 M
NaHCO3 at 65C for 15 minutes each. The eluates were pooled together and NaCl was
added to the eluates to a final concentration of 200 mM. Crosslinking was reversed by
incubation at 650C for 4 hours. RNA was digested by incubation with 40 pg/ml RNase
cocktail for 30 minutes at 370C. Proteins were then digested with 40 .g/ml proteinase K
in 10 mM EDTA, 40 mM Tris-HCl pH 6.5 at 370C for 1 hour. DNA was purified using a
purification kit (Qiagen) and eluted in 100 [l TE pH 7.4. 2.5 pl DNA was used as a
template for PCR with 50 pM primers and 10 pl PCR mix (Eppendorf) in a total volume
of 25 pl. Primer pairs were used against all areas of interest (Table 2-1). Primers for the
human y-globin genes were published by Schreiber et al .
Table 2-1. Names and sequences of primer pairs used for ChIP
Human HS2 5' flank
Mouse HS2 5'flank
US 5' GCTCCTTTATATGAGGCTTTCTTGG 3'
TATCTTAGAG GGAGGGCTGAGGGTTTG 3'
DS 5'AACATCTGGCCACACACCCTAAGC 3'
US 5' CTATTTGCTAACAGTCTGACAATAGAGTAG 3'
DS 5'GTTACATATGCAGCTAAAGCCACAAATC 3'
US 5'TTTACATAAGCCTAGTGGTACCCTTCC 3'
DS 5'ATCGCTGTCCTGCATCTCACAGTCG 3'
Titration of cycle numbers was performed with the different primer pairs to
determine the range of linear amplification. After the optimal number of PCR cycles,
DNA was run in 5% TBE polyacrylamide gels. The gels were stained with SyBr-green
and analyzed by fluorescence scanning on a Storm scanner. Quantitations were
performed using ImageQuant.
Chromatin Immunoprecipitation and DMS Footprinting
ChIP was performed as described above with the following modifications.
Approximately 3.33 x 107 MEL cells were collected per antibody to be used.
Crosslinking was induced by adding 1% (v/v) formaldehyde and incubation for 10
minutes at room temperature on a shaker. After stopping the crosslinking reaction by
adding 0.125 M glycine and incubation for 5 minutes (with shaking at room temperature),
the cells were pelleted at 2000 rpm. The cells were then washed twice in 25 ml ice-cold
phosphate-buffered saline (PBS) including protease inhibitors. Nuclei were isolated by
resuspending the cell pellet in 1 ml ice-cold swelling buffer (5 mM PIPES pH 8.0, 85 mM
KC1, 0.5% NP-40, and protease inhibitors), split into two aliquots and incubated on ice
for 10 minutes.
Chromatin was fragmented by subjecting the nuclei to restriction enzyme digestion
with 200 U PstI for 4 hours at 370C and 100 U PstI for an additional 16 hours at 370C.
The nuclei were then incubated with 200 URNase cocktail (Ambion) and an additional
100 U aliquot of Pstlfor 2 hours at 370C. Nuclei were pelleted at 40C for 5 minutes at
5000 rpm and lysed in 1 ml lysis buffer (1% SDS, 10 mMEDTA and 50 mM Tris-HCl
pH 8.1, protease inhibitors) on ice for 20 minutes. Restriction enzyme digestion was
verified by electrophoresis on a 2% agarose gel followed by Southern blotting using a
radioactive probe hybridizing to the human P-globin gene.
The lysate was combined and transferred to a 15 ml conical tube and diluted with 9
ml dilution buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl
pH 8.1, 167 mM NaCl and protease inhibitors). An aliquot of 500 [l protein A-
Sepharose beads (Amersham) was added to the diluted nuclear lysate and incubated for 2
hours at40C while rotating. The beads were pelleted for 10 minutes at 2000 rpm and the
supernatant was divided into three aliquots. An aliquot of 25 [il of the appropriate
antibody (USF 1 sc-229 or NF-E2 sc-291; Santa Cruz Biotechnology) or no antibody was
added to the aliquoted supernatant and incubated at 40C overnight while rotating. Protein
A-Sepharose beads were blocked overnight rotating at 40C in a 1:1 ratio of beads to
blocking buffer [3% BSA and 0.05% sodium azide in lx TE (10 mM Tris-HCl pH 8.1, 1
The chromatin was then immunoprecipitated with 600 pl blocked protein A-
Sepharose beads for 2 hours at 40C on a rotator. The immunoprecipitates were pelleted at
13000 rpm for 30 seconds and 1 ml of the no antibody supernatant was saved and labeled
as 'input'. Half of the input chromatin was ethanol precipitated and resuspended in two
aliquots of 20 pl ddH20 and 800 pl DMS buffer (50 mM sodium cacodylate, 1 mM
EDTA) and the other half was saved for the ChIP PCR analysis. The supernatants of the
samples precipitated with USF1 and NF-E2 antibodies were discarded and the pellets
were washed by rotating at 40C for 5 minutes with 1 ml each of low salt wash (0.1% SDS,
1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl pH 8.1, 150 mM NaC1), high salt wash
(0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl pH 8.1, 500 mM NaC1),
LiClwash (0.25 M LiC1, 1% NP-40, 1% sodium deoxycholate, 1 mMEDTA, 10 mM
Tris-HCl pH 8.1) and twice with lx TE pH 8. Of the immunoprecipitates, 80% was
resuspended in 800 pl DMS buffer and 20% was left in TE buffer for the ChIP/PCR
DMS treatment of the immunoprecipitated chromatin was performed using the
Maxam and Gilbert guanine-specific sequencing reaction with 0.1% DMS for 15, 45 or
90 seconds at room temperature . The reaction was stopped by adding 50 il DMS
stop buffer (1.5 M sodium acetate pH 7.0, 1 M 2-mercaptoethanol), followed by two
ethanol precipitations in a dry ice bath. The DMS-treated and non-DMS-treated
chromatin was then eluted from the beads by incubating twice with 250 Cl elution buffer
(1% SDS, 0.1 M NaHCO3), shaking at 950 rpm for 15 minutes at 650C, each time saving
the supernatant. An aliquot of 200 mM NaCl was added to the eluates and crosslinking
was reversed by incubation at 650C for 5 hours. Proteins were digested with 40 pg/ml
proteinase K in 10 mM EDTA and 40 mM Tris pH 6.5 for 1 hour at 370C.
Immunoprecipitated DNA was purified using a Qiagen kit and eluted with 180 pl ddH2O.
To cleave the DMS-treated DNA, 20 pl piperidine were added and incubated at 950C for
30 minutes. The DNA was washed twice by adding 1 ml ddH20, dried in a Speed Vac
and resuspended in 50 dl lx TE. Of the DMS-treated immunoprecipitated DNA, 10%
was used for ligation-mediated PCR (LMPCR)-assisted in vivo footprinting. An aliquot
of the precipitated DNA was also analyzed by PCR using primers specific for the murine
P-globin downstream promoter region (forward primer, 5'-
GACAAACATTATTCAGAGGGAGTACCC; reverse primer, 5'-
AGGTGCACCATGATGTCTGTTTCTGG) using a protocol previously published by
Forsberg et al .
LMPCR was essentially performed as described by Hornstra and Yang  with
the following modifications. Approximately 2 [g DMS-treated genomic DNA or 10%
immunoprecipitated DNA was annealed to 0.6 pmol gene-specific primer (MPA 1) by
denaturing at 960C for 10 minutes followed by annealing at 470C for 30 minutes in a 15
pl solution of lx Vent buffer (10 mM Tris-HCl pH 8.9 and 40 mM NaC1). Primer
extension was performed by adding a 15 pl solution of 10 mM Tris-HCl pH 8.9, 40 mM
NaC1, 0.5 mM dNTPs and 2 U Vent polymerase (New England Biolabs) and incubating
at 53C for 1 minute, 55C for 1 minute, 57C for 1 minute, 600C for 1 minute, 620C for
1 minute, 66C for 1 minute, 680C for 3 minutes and 76C for 3 minutes. A 20 il
dilution solution (110 mM Tris-HCl pH 7.5, 18 mM MgC12, 50 mM DTT, 0.0125%
BSA) was then added to the extension reaction.
Blunt-end ligation was performed by adding a 25 pl solution of 100 pmol
asymmetric double-stranded linker, 10 mM MgC12, 20 mM DTT, 3 mM ATP, 0.005%
BSA and 4.5 U T4 ligase (Ambion) to the reaction and incubating at 170C overnight. The
ligation products were purified by standard phenol/chloroform extractions and ethanol
precipitation including 10 pg/pl tRNA. The ligated DNA was resuspended in 20 pl
ddH20 and added to 80 il PCR mix [10 mM Tris-HCl pH 8.9, 40 mM NaC1, 3 mM
MgC12, 0.25 mM dNTPs, 20 pmol gene-specific PCRprimer (MPA 3), 20 pmol linker
primer 2 and 0.5 U Taq polymerase; Gibco BRL]. This PCR mixture was initially
denatured at 950C for 5 minutes and then subjected to 20 cycles of PCR under the
following conditions: 95C for 20 seconds, 65C for 1 minute, 720C for 1 minute with an
increase of 5 seconds/cycle and an additional 5 cycles of 95C for 20 seconds, 65C for 1
minute, 720C for 2 minutes 30 seconds, followed by a final extension at 720C for 15
minutes. The PCR products were purified by phenol/chloroform extraction and ethanol
precipitation and resuspended in 30 ul ddH20. Initially, 3 ll of the PCR products were
size-fractionated on a 0.4 mm thick 5% polyacrylamide gel made with 8 .g/ml
ammonium persulfate, then electrotransfered for 30 minutes to a nylon membrane
(Hybond N+; Amersham).
Radiolabeled probes were synthesized using the Prime-a-Probe kit (Ambion) from a
gel-purified PCR product (PCR primers: P-maj FWD, 5'-
GACAAACATTATTCAGAGGGAGTACCC-3'; MPB 1, 5'-
TCTGTCTCCAAGCACCCAA-3') containing the region of interest. To radiolabel the
probe, 150 ng template DNA was mixed with 0.3 pg gene-specific primer (MPA 3), used
in the PCR step of the LMPCR protocol, and brought up to 8 pl in ddH20. This primer
plus template mixture was denatured at 950C for 10 minutes and immediately placed on
dry ice to snap freeze. Then 5 il of 5x Decaprime buffer containing dATP, dGTP and
dTTP (Ambion) was incubated with 10 pl [a-32P] dCTP (3000 Ci/mmol) and 1 U/pl
Klenow fragment at 370C for 30 minutes. The reaction was quenched on ice and stopped
by adding 35 il formamide loading dye.
The probe was denatured at 950C for 10 minutes and purified on a 5% denaturing
polyacrylamide gel. After exposing the gel to film (Type 57; Polaroid) the probe was cut
out of the gel, crushed and soaked in 4 ml hybridization buffer (250 mMNa2PO4 pH 7.2,
7% SDS, 1% BSA). The probe was hybridized to the blots at 650C overnight. The blots
were washed three times at 650C in washing solution (20 mM Na2PO4 pH 7.2, 1% SDS)
and visualized by autoradiography. The following primers were used for LMPCR: MPA
1, 5'-ATGTCCAGGGAGAAATATCG-3'; MPA 3, 5'-
To 19 parts Laemmli sample buffer (Bio-Rad) was added 1 part P-mercaptoethanol.
2 volumes of this buffer were added to each nuclear protein extract or whole cell extract
and the mixture heated at 950C for 10 minutes. 25 [tl of each sample was loaded on a
10% Tris-HCl Ready Gel (Bio-Rad) and run at 150V in cold tank buffer (25 mM Tris, 192
mM glycine, 0.1% SDS) until the dye reached the bottom. The gel, membrane (Bio-Rad),
filter paper (Bio-Rad), and fiber pads (Bio-Rad Mini Protean apparatus) were soaked in
cold transfer buffer (10 mM NaHCO3, 3 mM NaCO3, and 20% methanol, pH 9.9) for 15
minutes and then set up in a transfer sandwich with a fiber pad, filter paper, the gel,
membrane, filter paper, and fiber pad. The transfer was run overnight at 30V at 40C with
constant stirring in the Mini Protean apparatus (Bio-Rad).
The transfer was disassembled and the membrane was stained with Fast Green for 5
minutes and washed twice in destain solution (50% methanol, 10% acetic acid). It was
then rinsed in TBS-Tween (30 mM Tris, 150 mM NaC1, 0.1% Tween 20, pH 7.6) and
blocked 1 hour in 5% milk in TBS-Tween shaking at room temperature. The membrane
was rinsed twice in TBS-Tween, washed once in TBS-Tween for 15 minutes while
shaking at room temperature, and washed twice for 5 minutes in TBS-Tween while
shaking at room temperature. It was incubated 1 hour in 10 ml 5% milk in TBS-Tween
containing 1:500 anti-HDAC3 while shaking at room temperature. The membrane was
rinsed twice in TBS-Tween, washed once in TBS-Tween for 15 minutes while shaking at
room temperature, and washed twice for 5 minutes in TBS-Tween while shaking at room
temperature. It was incubated 1 hour in 15 ml 5% milk in TBS-Tween containing
1:50000 anti-rabbit while shaking at room temperature. The membrane was rinsed twice
in TBS-Tween, washed once in TBS-Tween for 15 minutes while shaking at room
temperature, and washed twice for 5 minutes in TBS-Tween while shaking at room
temperature. ECL reagents (Amersham) were mixed according to the manufacturer's
recommendations and added to the membrane. The membrane was then exposed to
Kodak MS film for varying times.
K562 cells were synchronized at the border of G1 and S phase by incubating the
cells with complete growth medium in the presence of 2 mM thymidine for 17 hours,
washing twice with complete medium, incubating in complete medium for 12 hours and
blocking the cells in complete medium with 2 mM thymidine for another 17 hours [167,
168]. Blocked cells were fixed with ethanol, digested with RNase, stained with
propidium iodide and subjected to flow cytometry to verify synchronization. Blocked
cells were taken as "time 0" and subjected to ChIP and PCR analysis. The blocked cells
were then washed twice with complete medium to release them from the block and
allowed to grow in complete medium. Cells were taken for flow cytometry, ChIP, and
PCR analysis at specific time points after release from the arrest.
PCR-Based DNA Replication Analysis
Added to lysis solution (50 mM KC1, 10 mM Tris-HCl pH 8, 1.25 mM MgCl2,
0.45% NP-40, 0.45% Tween 20) was 75 [tl of 10mg/ml proteinase K per ml oflysis
solution (added fresh every time). Synchronized cells were counted and 106 cells were
collected, washed with lx PBS and resuspended in 100 [tl lysis solution per 105 cells.
The cells were incubated in the lysis solution overnight. The next day, 500 [tl phenol was
added and vortexed for 1 minute. The reaction was centrifuged for 5 minutes at 14000
rpm and room temperature. Then 500 [tl phenol/chloroform/isoamyl alcohol was added
to the supernatant and it was vortexed for 1 minute. The samples were centrifuged for 5
minutes at 14000 rpm and room temperature. Then, 500 [tl chloroform/isoamyl alcohol
was added to the supernatant, the reaction was vortexed for 1 minute and centrifuged for
5 minutes at 14000 rpm and room temperature. Next 100% ethanol was added to the
supernatant, incubated for 15 minutes at room temperature, and centrifuged 30 minutes at
14000 rpm and room temperature. The supernatant was discarded and 500 [tl of 70%
ethanol was added to the pellet. The sample was centrifuged for 10 minutes at 14000
rpm and room temperature. The supernatant was removed and the pellet allowed to air
dry. The pellet was then resuspended in 50 [tl TE pH 7.4. 1 [tl RNase was added and it
was incubated for 30 minutes in a 370C water bath. The DNA was used for PCR and the
rest stored at -200C.
2.5 pl DNA was used as a template for PCR with 50 pM primers and 10 tl PCR
mix (Eppendorf) in a total volume of 25 [il. Primer pairs were used against the human E-
globin, human necdin, mouse pmaj-globin and mouse necdin, as described in Table 2-1.
PCR was performed with cycle numbers 18, 20, 22, 24, 26, 28, and 30. DNA was run in
5% TBE polyacrylamide gels. The gels were stained with SyBr-green and analyzed by
fluorescence scanning on a Storm scanner. Quantitations were performed using
ImageQuant. The ratios of human e-globin to human necdin and mouse pmaj-globin to
mouse necdin were calculated for samples from all time points at a cycle number within
the linear range.
In Vitro Polymerase Transfer Analysis
A plasmid containing the wild-type human 3-globin LCR was linearized and
immobilized on streptavidin coated magnetic beads as described by Leach et al .
The immobilized LCR (200 ng) was incubated for 30 minutes at 300C with 50 tg MEL
nuclear protein extract (10 itl) and 15 [tl 2x binding buffer (36 mM HEPES, pH 7.9, 160
mM KC1, 40 mM MgC12, 4 mM DTT, 10 mM PMSF, 20% glycerol) in a total volume of
100 ptl. The tubes were placed on a magnet, the supernatant removed, and the beads were
washed three times with lx binding buffer. The beads were then resuspended in lx
binding buffer and incubated for 10 minutes at 300C in the presence or absence of 50 ng
of the plasmid pRS3A/X containing the wild-type or mutant P-globin gene . The
supernatants were removed and samples containing the wild type or mutant P-globin gene
were crosslinked by incubation for 10 minutes in 0.5 % formaldehyde at RT. The
crosslinking reaction was stopped by the addition of 0.125 M glycine and incubation for
5 minutes at RT. The supernatant of reactions not containing the P-globin gene were
incubated for 10 minutes at 300C in the presence of pRS3A/X, after which these samples
were crosslinked as well. All samples were dialyzed against ChIP dilution buffer and
subjected to immunoprecipitation using Pol II specific antibodies and PCR analysis as
described by Leach et al . In some experiments recombinant NF-E2  or BSA
(0.3 [tg/jtl) was added to the transfer reactions. The plasmid Pmaxi was used as a positive
control for the PCR and was generated by lighting an 800 bp Pmll restriction fragment
from the human LCR into the Pmll site of pRS 3A/X. The P-globin PCR primers span the
Pmll site present in the P-globin gene thus generating a PCR product that is 800 bp larger
than the PCR product from the wild type P-globin gene. The TK-HGH plasmid contains
the human growth hormone gene under the control of the Herpes Simplex Virus
thymidine kinase promoter . The HS2 plasmid contains the HS2 core plus flanking
sequence as previously described . The following primers were used in these
experiments: P-globin US: 5' CCTGAGGAGAAGTCTGCCGTTACTG 3' and DS: 5'
TCCTATGACATGAACTTAACCATAG 3'; TK-HGH: US: 5'
GGAGGCTGGAAGATGGCA 3'; DS: 5' AGTAGTGCGTCATCGTTGTGTG 3';
human HS2: as described in Table 2-1.
MEL Cell DMSO Differentiation
MEL cells were incubated in RPMI 1640 with L-glutamine supplemented with
10% fetal bovine serum, 1% antibiotic-antimycotic (ABAM) and 1% DMSO in 5% CO2
at 37C for three days. RNA was isolated from the cells for semi-quantitative RT-PCR.
RNA was isolated for RT-PCR using an RNA Isolation Kit (Gentra) according to
the manufacturer's protocol. Reverse Transcription was performed using 200 to 250 ng
RNA and the first strand cDNA synthesis Kit (Bio-Rad) as described by the
manufacturers protocol. PCR amplification was performed using PCR Mastermix from
Eppendorf and primer sequences specific for the murine P-globin and P-actin gene with
15 and 16 cycles as indicated. The mouse P-actin primers used were: US 5'
GGACGACATGGAGAAGAT 3' and DS 5' ATCTCCTGCTCGAAGTCT 3'. The
mouse P-globin primers used were: US 5' AAAGGTGAACTCCGATGAAGTTGG 3'
and DS 5' TTCTGGAAGGCAGCCTGTGC 3'. After electrophoresis the gels were
stained with SyBr Green, scanned using a storm scanner, and quantitated with Image
Approximately 2 x 107 K562 cells were collected per immunoprecipitate reaction.
The cells were incubated in 1% formaldehyde with rocking at room temperature for 10
minutes to induce crosslinking. Crosslinking was quenched with 0.125 M glycine for 5
minutes shaking at room temperature. Cells were washed twice with cold lx PBS with
Complete Protease Inhibitors (Roche), resuspended in swelling buffer (5 mM PIPES pH
8.0, 85 mM KC1, 0.5% NP-40, protease inhibitors), and incubated on ice for 10 minutes.
Nuclei were pelleted by centrifugation for 5 minutes at 5000 rpm and 40C. Nuclei were
lysed by resuspending in lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl pH 8.1,
protease inhibitors) and incubating on ice for 10 minutes. DNA was sonicated in an ice
water bath using a Fisher model 100 sonicator at power 5 with 5 pulses of 10 seconds
each with 1 minute cooling in between, to yield an average size of -500 bp. The sample
was then centrifuged for 10 minutes at 14000 rpm and 40C. The supernatant was diluted
in dilution buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 167 mM Tris-HCl
pH 8.1, 167 mM NaC1, protease inhibitors) to a final volume ofx ml where x = # of
antibody samples needed. Then 50 pl of Protein A-Sepharose beads (Amersham) per 2 x
107 cells used were added to the diluted lysate and incubated for 2 hours with rotation at
4C. The beads were pelleted and 5 il of appropriate antibody was added to each ml of
aliquoted supernatant and incubated overnight with rotation at 40C. The antibodies used
TFII-I (a gift from R. Roeder, Rockefeller University) and HDAC3 (a gift from E. Seto,
University of South Florida). Protein A-Sepharose beads were blocked by resuspension
in blocking buffer (3% BSA, 0.05% Na azide in 10 mM Tris-HC1 pH 8.1, 1 mM EDTA)
and incubated overnight with rotation at 40C.
The next day, the samples were rotated for 2 hours at 40C with 60 pl blocked
Protein A-Sepharose beads. The beads were pelleted and supernatants removed. Then
500 pl of the supernatant of the 'no antibody' sample was kept and labeled 'input'. The
pelleted beads were washed with low salt wash (0.1% SDS, 1% Triton X-100, 2 mM
EDTA, 200 mM Tris-HCl pH 8.1, 150 mM NaC1), high salt wash (0.1% SDS 1% Triton
X-100, 2 mM EDTA, 200 mM Tris-HC1 pH 8.1, 500 mM NaC1), LiCl wash (0.25 M
LiC1, 1% NP-40, 1% Nadeoxycholate, 1 mM EDTA, 100 mM Tris-HCl pH 8.1) and two
washes with TE pH 8. DNA was eluted off the beads with two 250 ll 1% SDS and 0.1 M
NaHCO3 washes at 650C for 15 minutes each shaking at 950 rpm. The two elutions were
pooled together and then dialyzed against 200 ml of dilution buffer with Complete Mini
protease inhibitors using Slide-A-Lyzer Mini Dialysis units (Pierce) at 40C with constant
stirring for 2 hours. After samples were taken out of dialysis units and placed in 1.5 ml
tubes, 500 [tl dilution buffer was added to each. For the second immunoprecipitation 5 dtl
of antibody was added to each and incubated overnight at 40C rotating. Protein A-
Sepharose beads were blocked by resuspending them in blocking buffer and incubated
overnight with rotation at 40C.
The next day, the samples were rotated for 2 hours at 40C with 60 pl blocked
Protein A-Sepharose beads. The beads were pelleted and supernatants removed. The
pelleted beads were washed with low salt wash, high salt wash, LiCl wash and two
washes with TE pH 8. DNA was eluted off the beads with two 250 pll 1% SDS and 0.1 M
NaHCO3 washes shaking at 650C for 15 minutes each. The two elutions were pooled
together and NaC1 was added to the eluates to a final concentration of 200 mM and
crosslinking was reversed by incubation at 650C for 4 hours. RNA was digested by
incubation with 40 pg/ml RNase cocktail (Ambion) for 30 minutes at 370C. Proteins
were then digested with 40 pg/ml proteinase K in 10 mM EDTA, 40 mM Tris-HCl pH
6.5 at 370C for 1 hour. DNA was purified using a column purification kit (Qiagen) and
eluted in 100 pl TE pH 7.4. Then 2.5 pl DNA was used as a template for PCR with 50
pM primers and 10 pl PCR mix (Eppendorf) in a total volume of 25 il. The human 3-
globin, e-globin and HS2&3 linker primer pairs were used for PCR (see Table 2-1). PCR
was performed for 28 cycles, and DNA was run in a 5% TBE polyacrylamide gel. The
gels were stained with SyBr-green and analyzed by fluorescence scanning on a Storm
107 K562 cells were collected per immunoprecipitate reaction. Cells were washed
twice with cold lx PBS with Complete protease inhibitors (Roche), resuspended in
swelling buffer (5 mM PIPES pH 8.0, 85 mM KC1, 0.5% NP-40, protease inhibitors) and
incubated on ice for 10 minutes. Nuclei were pelleted by centrifugation for 5 minutes at
5000 rpm and 40C. Nuclei were lysed by resuspending in lysis buffer (1% SDS, 10 mM
EDTA, 50 mM Tris-HCl pH 8.1, protease inhibitors) and incubating on ice for 10
minutes. DNA was sonicated in an ice water bath with a Fisher model 100 sonicator at
power 5 with 5 pulses of 10 seconds each with 1 minute cooling in between, to yield an
average size of -500 bp. The samples were then centrifuged for 10 minutes at 14000 rpm
and 40C. The supernatant was diluted in dilution buffer (0.01% SDS, 1.1% Triton X-100,
1.2 mM EDTA, 167 mM Tris-HCl pH 8.1, 167 mM NaC1, protease inhibitors) to a final
volume of x ml where x = # of antibody samples needed. 50 pl of Protein A-Sepharose
beads (Amersham) per 107 cells used were added to the diluted lysate and incubated for 2
hours with rotation at 40C. The beads were pelleted and 5 pl of appropriate antibody was
added to the aliquoted supernatant and incubated overnight with rotation at 40C. The
antibodies used were USF1 sc-229, USF2 sc-862, Pol II sc-899 (all from Santa Cruz)
TFII-I (a gift from R. Roeder, Rockefeller University) and HDAC3 (a gift from E. Seto,
University of South Florida). Protein A-Sepharose beads were blocked by resuspension
in blocking buffer (3% BSA, 0.05% Na azide in 10 mM Tris-HCl pH 8.1, 1 mM EDTA)
and incubation overnight with rotation at 40C.
The next day, the samples were rotated for 2 hours at 40C with 60 pl blocked
Protein A-Sepharose beads. The beads were pelleted and supernatants removed. The
pelleted beads were washed with low salt wash (0.1% SDS, 1% Triton X-100, 2 mM
EDTA, 200 mM Tris-HCl pH 8.1, 150 mM NaC1), high salt wash (0.1% SDS 1% Triton
X-100, 2 mM EDTA, 200 mM Tris-HC1 pH 8.1, 500 mM NaC1), LiCl wash (0.25 M
LiC1, 1% NP-40, 1% Nadeoxycholate, 1 mM EDTA, 100 mM Tris-HCl pH 8.1) and two
washes with TE pH 8. DNA was eluted off the beads with two 250 pl 1% SDS and 0.1 M
NaHCO3 washes shaking at 650C for 15 minutes each. The two elutions were pooled
together and concentrated in the Speed Vac for about 1 hour. The samples were then
subjected to Western blotting procedures as described above.
CHROMATIN IMMUNOPRECIPITATION AND DMS FOOTPRINTING
Transcription in eukaryotes is a complex process involving the binding of proteins
to the promoter, recruitment of transcription complexes, initiation, elongation and
termination . This process is dynamic and controlled at each step by proteins that
bind to the DNA, to the RNA polymerase holocomplex, or both. Although knowledge
about transcription complex formation in vitro is extensive, the mechanisms leading to
transcription in the context of higher order chromatin in vivo are not understood in detail.
Many nuclear processes including transcription, DNA replication, homologous
recombination, DNA repair, and others are regulated by proteins that interact with
specific DNA sequences or structures in the context of chromatin. Therefore, there is a
great demand for new and better techniques to investigate protein-DNA interactions. A
number of techniques are available to study the interaction of proteins with specific
segments of DNA in vivo; each of these techniques has distinct limitations [166, 171-
Chromatin immunoprecipitation (ChIP) is a powerful method to analyze the
interaction of proteins with specific chromatin regions in vivo [171, 172]. This technique
involves crosslinking protein-DNA and protein-protein interactions in live cells, followed
by immunoprecipitation of the lysed nuclear contents to isolate segments of DNA that are
bound to the protein of interest in vivo. Combined with PCR, the ChIP assay is very
sensitive in detecting proteins that crosslink to a specific region in chromatin. There are
at least two limitations of the ChIP assay. First, ChIP does not provide information
regarding the specific sequence a protein interacts with in vivo. The size of the ChIP
fragments is on average 500bp, so the corresponding binding site for a specific protein
can be within a region of up to 1 kb of DNA. Second, ChIP does not reveal whether a
protein directly binds to the DNA or is recruited to DNA by protein/protein interactions.
Another method commonly used to examine the interaction of proteins and DNA is
in vivo footprinting [166, 173, 174]. In footprinting experiments enzymes like DNaseI or
chemicals like DMS/piperidine are used to cleave the DNA. Nucleotides that are bound
by proteins are normally protected from cleavage. The cleavage pattern can be visualized
either directly or after PCR using electrophoresis in combination with radioactive
labeling techniques and autoradiography. The limitation of in vivo footprinting is that it
does not reveal what the protein is that is causing the protection of particular bases.
Another problem with the in vivo footprinting technique is that it fails to provide clear
results in situations where a protein is bound in a sequence-specific manner in only a
small fraction of cells, in other words, if the protein-DNA interaction is transient.
We have combined the ChIP and in vivo footprinting techniques and developed a
novel method for analyzing protein-DNA interactions in vivo. This new technique is
similar in concept to an in vitro protocol developed by Gallarda et al. . These
authors used a monoclonal antibody-based DNaseI footprint selection technique, which
unambiguously identifies proteins responsible for particular footprints.
The technique we have developed here is aimed at analyzing sequence-specific
interactions of particular proteins in vivo. In the process of developing and optimizing the
technique, we have analyzed the interaction of proteins with the mouse P-globin
downstream promoter region in unsynchronized MEL (mouse erythroleukemia) cells.
We show that chromatin fragments precipitated with antibodies against USF 1 or NF-E2
reveal strikingly different footprint patterns on a specific P-globin gene promoter region.
The USF 1-selected templates show no footprint in this region, but the NF-E2-selected
fragments show a footprint at a non-consensus MARE. This result indicates that this
novel method will allow investigators to select specific templates for in vivo footprinting,
and determine exactly where their protein of interest binds.
The developmental stage-specific expression of the human P-globin gene is
regulated primarily by transcription factors that interact with gene-proximal cis DNA
elements . We have shown that USF1 and USF2 interact with a downstream promoter
element of the human 3-globin gene in vitro . In the same study we have shown that
these proteins are crosslinked to the P-globin gene in erythroid cells in vivo. In addition,
data published by Sawado et al. , as well as our own experiments revealed that NF-
E2 can be crosslinked to the P-globin gene in vivo. The crosslinking of NF-E2 to the
globin promoter is somewhat surprising as there is no consensus DNA-binding site in the
globin gene. However, a sequence located downstream of the transcription start site
reveals partial homology to the NF-E2 consensus element, also referred to as the MARE
(maf recognition element, Figure 3-1; [126, 177, 178]. The NF-E2 consensus sequence is
5'-TGCTGASTCAY-3' (S = G or C; Y = T or C; ). The sequence in the mouse P-
globin downstream promoter region differs in one position (the third) from the consensus
sequence. Conventional in vivo footprinting did not reveal significant protection of the
NF-E2-binding site in MEL cells (Figure 3-4, lanes 1 and 2). To analyze the possibility
that NF-E2 interacts with the P-globin downstream element in vivo in only a fraction of
erythroid cells, we have developed a novel method that combines the ChIP assay with
Initiasor MARE/APl-like element
R: ctgcttgeacttgcttttgacacaact tgtttacttgcaatccccccaaaac
Figure 3-1. Sequence alignment of the adult P-globin downstream promoter region.
Shown are three sequences of the P-globin downstream promoter region
from human (H), mouse (M) and rabbit (R) . The shaded box
highlights the position of the MARE/Ap -like sequence and the arrow
points to the transcription start site.
The procedure (outlined in Figure 3-2) involves the crosslinking of proteins and
DNA using formaldehyde (1%), followed by digestion of permeabilized nuclei with
restriction enzymes, precipitation with specific antibodies, treatment and cleavage of
precipitated chromatin restriction fragments with DMS and piperidine and analysis of the
cleavage pattern by LMPCR (for details see Materials and Methods Chapter).
Figure 3-3 shows that antibodies against USF1 and NF-E2 precipitate crosslinked
chromatin fragments containing the murine adult P-globin gene, as expected from
previous data [163, 176].
The subsequentDMS footprint of the precipitated chromatin fragments revealed
interesting characteristics of the precipitated fragments (Figure 3-4). First, the overall
footprint pattern of fragments selected with USF antibodies is different from those
precipitated withNF-E2 antibodies (compare Figure 3-4, lanes 3 and 4). This is
particularly obvious over the MARE-like sequence, which shows clear protection only in
templates selected by NF-E2 precipitation.
Incubale with restriction enzyme X
Precipitate with antibody
Specific for protein A Y
Treat with DMS
Treat with Piperidene
Figure 3-2. Analysis of protein-DNA interactions in the murine P-globin downstream
promoter region by a combination of chromatin immunoprecipitation and
DMS footprinting. Diagram outlining the experimental procedure for
footprinting ChIP-selected templates.
The difference in the cleavage pattern between the templates selected by USF or
NF-E2 antibodies could reflect the possibility that the proteins interact with the globin
gene at different stages of the transcription cycle. Alternatively, the results could reflect
the possibility that low occupancy of these sites would strongly reduce the probability of
selecting templates which have both proteins bound simultaneously. Second, the overall
footprint pattern of fragments precipitated with USF antibodies is similar to the pattern
found in unselected templates, whereas the footprint pattern for NF-E2-selected
fragments is different. This result suggests that only a small fraction of cells may have
NF-E2 bound at the 3-globin gene.
1 2 3 4
Figure 3-3. ChIP experiment showing the interaction of USF 1 and NF-E2 with the
murine pmaj-globin promoter in MEL cells. MEL cells were grown under
standard conditions and crosslinked with formaldehyde. After
fragmentation, the chromatin was precipitated with no antibodies (lane 2),
antibodies against USF 1 (u.USF, lane 3) or antibodies against the p45
subunit of NF-E2 (ucNF-E2, lane 4). Lane 1 shows the PCR result of the
input, which serves as a positive control.
It is important to note that formaldehyde crosslinking does not appear to change the
cleavage pattern in the DNA region analyzed in these experiments; the overall pattern is
similar between untreated cells (Figure 3-4, lanes 1 and 2) and crosslinked cells (lanes 3
and 4). In addition, we found that incubation of genomic DNA with 1% formaldehyde
(with subsequent reversal of the crosslink) does not change the overall DMS cleavage
pattern in the DNA region that we have analyzed by LMPCR (data not shown).
We examined the interaction of NF-E2 with the non-consensus MARE-binding site
in the downstream promoter region in detail using a novel in vivo method. The potential
sequence-specific interaction of NF-E2 with the 3-globin downstream promoter region is
interesting in light of the fact that Johnson et al.  previously showed that RNA
polymerase II is recruited to both the LCR and to the adult 3-globin gene and that the
transfer of the polymerase from the LCR to the globin gene depends on the presence of
NF-E2 (p45). In this respect it is possible that NF-E2 is part of the recruitment process
that mediates the interaction of RNA polymerase II with the P-globin promoter.
i .i A T
1 2 3 4 5
LMPCR footprint analysis of non-selected or ChIP-selected chromatin.
Chromatin was precipitated with USF or NF-E2 antibodies. The precipitates
were subjected to DMS footprinting (lane 3, USF-selected chromatin; lane
4, NF-E2-selected chromatin). In vivo footprinting was also performed on
MEL cells that were not treated with formaldehyde (lane 2). Lane 1 shows
the DMS/piperidine cleavage pattern in in vitro treated MEL genomic DNA.
Lane 5 shows the G-ladder of the P-globin downstream promoter region.
Indicated on the right are the positions of the initiator and MARE-like
sequences, as well as the start site and direction of transcription. Open
circles indicate protected G residues in NF-E2-selected chromatin.
Several methods are available to fragment crosslinked chromatin for ChIP,
including sonication, MNase digestion and restriction enzyme digestion [171, 172, 180,
181]. We reasoned that sonication or MNase digestion might fragment the chromatin in a
way that not all of the resulting templates would be amplifiable by LMPCR and in
addition may result in a high background. We therefore chose to fragment the
crosslinked chromatin by restriction enzyme digestion . The disadvantage in using
this method is that long incubation with the restriction enzyme allows endogenous
nucleases to attack the chromatin in accessible regions. In some of our experiments, we
observed inconsistent LMPCR banding patterns in ChIP-selected templates in the higher
molecular weight ranges (data not shown). This problem can be solved by incubating the
nuclei for shorter periods of time with higher concentrations of restriction enzyme 
or by following similar restriction digest protocols from other groups .
Another issue that has to be considered is the fact that the precipitation of
templates using ChIP results in different concentrations of templates. It is thus important
to perform a titration of the templates precipitated with different antibodies. Similarly,
we were unsure whether the steps of the ChIP assay leave the protein in a native structure
to give a true footprint in all cases tested. Though it could be postulated, that once a
footprint such as NF-E2 at an NF-E2 binding site has been observed that a similar
treatment will leave other proteins in a conformation that makes a footprint detectable.
Lastly, it was suggested that the antibody might remain bound and contribute to the
footprint observed. Therefore, to remove this possibility it would be possible to change
the elution buffer to one that is known for disrupting the antibody-antigen interactions,
such as the ImmunoPure Gentle Ag/Ab Elution Buffer available from Pierce
INTERACTION OF RNA POLYMERASE II AND TRANSCRIPTION FACTORS
WITH THE P-GLOBIN LOCUS
The five genes of the human 3-globin locus are expressed in erythroid cells in a
tissue- and developmental-stage specific manner . Appropriate expression of the
globin genes is regulated by many DNA elements that are located proximal or distal to
the genes. The human P-globin locus control region (LCR) is a powerful regulatory
DNA element located far upstream of the genes and required for high-level expression of
all the globin genes throughout development [136, 183]. The LCR, unlike classical
enhancer elements, operates in an orientation dependent manner . There is currently
no consensus on how the LCR acts to stimulate globin gene transcription but it is
generally believed that it involves some form of communication between the LCR and
the globin genes [101, 115, 184]. The LCR is composed of several regions that exhibit
extremely high sensitivity to DNaseI in erythroid cells (hypersensitive HS sites 1 to 5).
The core HS sites contain clusters of transcription factor binding sites and are separated
from each other by 2 to 4 kbp of DNA . Results from analyzing human LCR
function at ectopic sites in the context of transgenic mice suggests that the HS sites
synergistically enhance globin gene transcription [19, 20, 119, 121-123], whereas studies
in the endogenous murine locus show that the core HS sites function additively [109, 185,
Recent models view the LCR as a holocomplex in which the individual HS sites
interact via extensive protein-DNA and protein-protein interactions [118, 119, 134]. The
LCR holocomplex may provide a highly accessible region for the efficient recruitment of
macromolecular complexes involved in chromatin modification and transcription .
Indeed, it has been shown that RNA polymerase II (Pol II) transcription complexes are
recruited to LCR HS sites in vitro and in vivo [169, 187-189], suggesting that
transcription complexes are first recruited to the LCR and subsequently transferred to the
globin genes . Sawado et al.  recently demonstrated that another important
function of the LCR is to regulate transcription elongation at the adult P-globin gene.
An interesting aspect of gene regulation is the question of how transcriptional
activity is maintained during the process of replication . Activation of at least some
gene loci depends on replication, and it is possible that replication could provide a
window of opportunity for transcription factors to gain access to regulatory sequences
before repressive chromatin is formed.
Generally, transcriptionally active regions of the genome replicate early, whereas
repressed parts of the genome replicate late. A recent study showed that plasmids
injected into cells at different time points during replication adopt an open chromatin
conformation at earlier time points and a repressed conformation at later time points
suggesting that early replication favors the assembly of accessible chromatin, or that
accessible chromatin is replicated early . Accordingly, the globin locus replicates
late in non-erythroid cells and early in cells expressing the globin genes. Transgenic
studies have shown that the timing of replication is regulated by the LCR , however,
this activity, like the chromatin opening activity, appears to be redundant in the
endogenous P-globin locus . It is not known how or if replication affects the
association of transcription factors and Pol II with the globin locus.
Here, we analyzed the interaction of transcription complexes with the human and
murine P-globin loci in erythroid cells expressing either the human embryonic/fetal e-
and y-globin (K562 cells) or the murine adult P-globin genes (MEL cells) using
chromatin immunoprecipitation (ChIP).
We began our studies by examining the interaction of various components of
transcription complexes and erythroid transcription factors with the P-globin gene locus
in murine and human erythroleukemia cell lines. Figure 4-1 shows a diagram of the
human and murine P-globin loci and the localization of primers used in the ChIP
We used chromatin immunoprecipitation (ChIP) and PCR to analyze the interaction
of RNA Pol II (Pol II), TBP, TFIIB, NF-E2 (p45), and acetylated histone H3 with
different regions of the P-globin gene locus. As a negative control we also analyzed the
interaction of the proteins with the necdin gene, which is not active in erythroid cells.
NF-E2 is a hematopoietic specific transcription factor composed of a large (p45) and a
small (p 18) subunit  and known to interact with several sequences in the human and
murine P-globin loci [136, 195]. We investigated the interaction of these proteins in two
cell lines; a human erythroleukemia cell line (K562) expressing the embryonic s and fetal
y-globin genes but not the adult P-globin gene, and a murine erythroleukemia cell line
(MEL), in which the adult Pmaj- and pmin-globin genes are expressed but not the
embryonic sy and 3hl genes.
Human p-globin gene locus
LCR HS 5 4 3 2 1
E Gy Ay
hHS2 5'flank hHS2 he hy hy hp
Mouse p-globin gene locus
LCR HS 5 4 3 2 1 sy ph1 pmaj pmin
mHS2 5'flank mHS2 mEy mpmaj
Diagrammatic representation of the human and murine P-globin gene locus.
The human 3-globin gene locus: the embryonic s-globin gene, the two fetal
y-globin genes, and the adult 6- and P-globin genes. The LCR is located
upstream of the s-globin gene and composed of at least 5 HS sites. The
murine P-globin locus: the sy and 3hl globin genes, which are co-expressed
in the embryonic yolk sac and the pmaj and 3min globin genes, which are
expressed during the fetal and adult stages of erythropoiesis. The overall
organization of the LCR is highly conserved between the murine and the
human 3-globin locus. PCR fragments analyzed in the ChIP experiments
are indicated as lines (horizontal bars) below the respective regions.
The data show that RNA polymerase II is efficiently crosslinked to LCR HS2 in
K562 cells (Figure 4-2). The interaction of Pol II with the globin genes is consistent with
the expression pattern in these cells. Pol II is detectable at the embryonic s- but not at the
adult P-globin gene. The results further show that Pol II does not interact with the HS2
5'flanking region, demonstrating that Pol II specifically interacts with the core of HS2.
We also detected interactions of phosphorylated Pol II with HS2 and the s-globin gene in
K562 cells, indicating ongoing transcription in these regions.
Because extragenic transcripts have been detected over the entire LCR in the
human P-globin locus we analyzed the interaction of basal transcription factors that are
not associated with the transcription elongation complex. The results show that both
TFIIB as well as TBP can be crosslinked to HS2 in K562 cells. These results
demonstrate that Pol II and other basal transcription factors, which are part of
transcription initiation complexes, are associated with the LCR element HS2 in vivo. The
presence of components of the PIC indicates that the crosslinking of Pol II to HS2 is not
solely due to elongating polymerase. NF-E2 is clearly detectable at HS2 in K562 cells,
whereas the signal representing the e-globin gene is very weak. None of the proteins
interact with the necdin gene.
Figure 4-2. Interaction of transcription factors and RNA polymerase II with the human
P-globin locus and the human necdin gene. Antibodies used were specific
for RNA polymerase II (aPol II), RNA polymerase II phosphorylated at the
C-terminal domain (aPol II Phosph.), TBP, TFIIB, NF-E2 p45, and acetyl
histone H3 (ahistone H3 acet.). PCR was performed with primers to
regions in the P-globin locus as indicated (see Figure 4-1). No antibody
samples are negative controls. Input represents the positive PCR control
taken from the no antibody sample.
The observation that Pol II is not detectable at the HS2 5'flanking region is
interesting and suggests that transcription initiates within the HS2 core enhancer and not
upstream of HS2. In order to analyze this question in more detail, Padraic Levings in the
laboratory used nuclear run-on experiments to determine regions of active transcription in
the human P-globin locus in K562 cells. He found that ongoing transcription is
detectable in a region upstream of HS5, in the HS2 core, and in the ; globin gene, but not
in the HS2 5' flanking region. These results are consistent with our ChIP analysis and
suggest that transcripts initiating far upstream of HS2 do not elongate beyond the HS2
5'flanking region analyzed here, consistent with data published by Kim and Dean .
Because the nuclear run-on HS2 probe is specific for the core region his data suggest that
transcription initiates within HS2 and not, or not exclusively, downstream of HS2 as
proposed by Routledge et al. . Transcription upstream of HS5 probably initiates
within the LTR of a retrovirus as previously described .
We also analyzed the interaction of transcription factors and Pol II with regulatory
elements in the murine P-globin locus in MEL cells (Figure 4-3A). The results of these
experiments are very similar to those in the K562 cells and show that Pol II, TBP, and
TFIIB can be crosslinked to both HS2 and the P-globin gene promoter, but not to a region
5' of HS2. The precipitation of proteins crosslinked to the e-globin gene is extremely
inefficient. The inefficiency in crosslinking phosphorylated Pol II to HS2 and the 3-
globin promoter in the murine locus is consistent and may be due to species-specific
differences in the ability of the antibody to precipitate crosslinked polymerase.
In contrast to the results in K562 cells, NF-E2 (p45) can be crosslinked to HS2 and
to the P-globin gene promoter. We have shown previously that NF-E2 interacts with a
NF-E2 like sequence located downstream of the murine pmaj-globin transcription start
site . This sequence deviates in one position from what is considered the consensus
NF-E2 binding site, but it is identical to the NF-E2 interaction sequence present in human
LCRHS3 [126, 136].
g fl%'g'5 ii
Interaction of transcription factors with and transcription in the murine 3-
globin locus in MEL cells. A) Interaction of transcription factors and RNA
polymerase II with the murine P-globin locus and the murine necdin gene.
Antibodies used were specific for RNA polymerase II (cPol II), RNA
polymerase II phosphorylated at the C-terminal domain (cPol II-Ph), TBP,
TFIIB, NF-E2 p45, and acetylated histone H3 (cAcetyl-H3). PCR was
carried out with primers corresponding to regions in the murine P-globin
locus as indicated (see Figure4-1). The no antibody and input controls were
processed as described in the legend to Figure 4-2. B) Semi-quantitative RT-
PCR analysis of P-globin gene transcription in DMSO induced and
uninduced MEL cells. Relative expression of P-globin mRNA compared to
expression of P-actin was found to be 3-fold lower in uninduced MEL cells.
The NRO analysis in MEL cells performed by Padraic Levings demonstrates that
while HS2 and the P-globin gene are transcribed, a region upstream of HS5 is not. This
result is likely due to the fact that the retrovirus element present upstream of HS5 in the
human locus is not present in the murine P-globin gene locus.
The MEL cells employed in these experiments express the P-globin gene, and thus
no induction (with DMSO) was used. Moreover, interactions between Pol II and NF-E2
with the globin locus can be detected by ChIP assay in the absence of DMSO. We found
that incubating the cells with 1% DMSO for 3 days led to an increase in globin gene
transcription (ca- and P-globin) less than 5-fold compared to uninduced cells (Figure 4-3
If the model according to which Pol II is recruited first to the LCR, modified and
transferred to the globin genes is correct , one may predict that Pol II associates with
the LCR and the globin genes at different time points during activation of the globin
locus in the cell-cycle. To address this question, we examined the interaction of
transcription complexes with the LCR and globin genes during S phase in K562 cells
(Figure 4-4). Cells were synchronized using the double thymidine block method, which
arrests the cells at the G1/S phase border . At specific time points after release from
cell-cycle arrest, cells were subjected to formaldehyde crosslinking and ChIP analysis.
In these experiments, we used antibodies against Pol II, NF-E2 (p45) and USF2.
We have shown that USF2 interacts with the human 3-globin promoter in K562 cells
(described in Chapter 6) and hypothesized that it may contribute to the repression of the
P-globin gene . We analyzed the association of these proteins with four different
regions of the human P-globin locus. Cells were crosslinked at time 0, and 15 minutes,
45 minutes, 2 hours, and 6 hours after release from cell cycle arrest.
At time 0, Pol II interacts with HS2, the e-, and y-globin genes, but not with the 3-
globin gene. There is no interaction of NF-E2 or USF2 with the regions of the globin
locus at this time point. The only change after 15 minutes into S phase is the association
of NF-E2 with HS2. After 45 minutes, Pol II is no longer detectable at the e-globin gene
but remains associated with HS2 and the y-globin genes. At this time point, NF-E2
interacts with the y-globin genes.
Ui ychm airte d
D N AC.n"nt'
Time 6- 5
0 inin -
45 min -
2hrs E- ^_
n = ,, w ,
, u -----
I M *-
Figure 4-4. Interactions of transcription factors and RNA polymerase II with the human
P-globin locus during early S phase in synchronized K562 cells.
Synchronized K562 cells were analyzed for DNA content by flow cytometry
(shown on top). The remaining cells were harvested at time 0 or at specific
time points after release from cell cycle arrest (as indicated), and subjected
to ChIP using antibodies specific for RNA polymerase II (cPol II), the p45
subunit ofNF-E2 (cNF-E2), and USF2 (cUSF2); time 0 represents the
status of the blocked cells at the G1/S phase boundary. PCR was performed
using primers specific for HS2, the e-globin gene, the y-globin gene, and the
P-globin gene as indicated (see Figure 4-1). The no antibody and input
controls were processed as described in the legend to Figure 4-2.
After 2 hours, Pol II no longer interacts with the LCR or the e- and y-globin genes.
At this time there is a clear interaction of NF-E2 and USF2 with HS2 and the e- and y-
globin genes. The interaction of NF-E2 with the P-globin gene is less pronounced. After
6 hrs, Pol II re-associates with HS2 and the y-globin genes, but not with the e-globin
gene. Accompanied with the reappearance of Pol II is the dissociation of USF2 from the
genes and HS2.
We also analyzed the interaction of transcription factors with the murine globin
locus during S phase in synchronized MEL cells (Figure 4-5). We used antibodies
specific for Pol II, TFIIB, NF-E2, and acetylated histone H3. At time zero, Pol II, TFIIB,
NF-E2 and acetylated histone H3 can be detected at LCR HS2 and the P-globin promoter.
After 7 minutes, NF-E2 and acetylated H3 but not Pol II can be detected at HS2 and
weaker at the P-globin gene. A similar pattern was observed after 45 minutes, except that
now Pol II is beginning to re-associate with HS2 and the P-globin promoter. Finally,
after 2 hours, Pol II, TFIIB, NF-E2, and acetylated H3 are associated with HS2 and the 3-
The results show that Pol II dissociates and re-associates with the globin locus
during early S phase in MEL cells. In contrast to the results with K562 cells, NF-E2
appears to remain associated with the globin locus at all the time points analyzed here.
All experiments have been repeated several times with essentially the same results. The
only exception is that in some experiments NF-E2 (p45) can be crosslinked to HS2 in
synchronized K562 cells at time 0. The variation may be due to subtle changes in the
status of the cells during synchronization.
F L& Unsync.
w] h 7rin
S P, E i d
Interactions of transcription factors and RNA polymerase II with the human
P-globin locus during early S phase in synchronized MEL cells. The cells
were analyzed for DNA content by flow cytometry at each time point
(shown at left). The remaining cells were harvested at time 0 or at specific
time points after release from cell cycle arrest (as indicated), and subjected
to ChIP using antibodies specific for RNA polymerase II (cPol II), TBP
(aTBP), TFIIB (cTFIIB), the p45 subunit of NF-E2 (cNF-E2), and
acetylated histone H3 (aH3acetylated); time 0 represents the status of the
blocked cells at the G1/S phase boundary. PCR was performed using
primers specific for HS2 and the pmaj-globin gene as indicated (see Figure
4-1). The no antibody and input controls were processed as described in the
legend to Figure 4-2.
We next analyzed how dissociation and re-association of Pol II to the P-globin
locus in K562 and MEL cells correlates with the timing of replication. Since the cells are
blocked at the border of G1 and S phase, it is assumed that when the cells are released
from the block they will synchronously enter S phase. They should then undergo
replication early in S phase since the P-globin locus is early replicating in erythroid cells
such as K562 and MEL . As a late replicating control, we used the necdin gene,
which is a brain-specific imprinted gene that replicates late in S phase in all tissues
In this experiment, the cells were blocked at the border of G1 and S phase, cells
collected at time 0, released from the block and cells collected at the same time points as
for the flow cytometry and ChIP experiments. DNA was extracted and used as a
template for PCR with primers against e-globin and necdin for K562 cells (Figure 4-6),
P-globin and necdin for MEL cells (Figure 4-7). Based on the calculated ratios it was
estimated that K562 cells complete replication of the P-globin locus by 2 hours after
release from the block (Figure 4-6). For MEL cells it was estimated that replication of
the P-globin locus is also complete by 2 hours after release from the block (Figure 4-7).
cycle# it # Ratio of /Necdin at cycle 28
-n Time 0 =0.32
,15 min = 0.42
S45 in = 0.52
2 hr = 0.86
-6 hr = 0.87
Figure 4-6. Semi-quantitative PCR analysis of DNA content of synchronized K562 cells.
Cells were isolated at each time point, DNA extracted and PCR performed
with primers to e-globin and necdin promoters. Equal amounts of the two
PCR products were run in the same lane and gels were stained with SyBr
Green. Bands were quantitated with Image Quant and the ratio of e-globin to
necdin was calculated.
Expression of the P-globin genes is regulated by the LCR [183, 199]. Whereas
studies using transgenic mice demonstrate that the LCR has both chromatin opening and
enhancer activities , studies in the endogenous murine locus show that the function
of the LCR is limited to stimulating high-level globin gene transcription . The
ability of the LCR to regulate chromatin structure is thought to be important for
conferring protection from position effects at ectopic sites in transgenic studies, a
defining activity for locus control regions . It is not known at what level the LCR
opens chromatin structure, e.g., unfolding of higher order structure and modification of
histones, or whether chromatin opening is a secondary event and the consequence of LCR
mediated re-location of the globin locus into a transcriptionally active site in the nucleus
DNA Analysis Using Ratio of P3/Necdin in MEL
Time 0 7 Min 45 Min 2 Hr
Cycle # 20 22 2426 20 22 2426 2022 2426 20222426
P/Necdin 1.27 1.29 1.68 2.68
at cycle 20
Figure 4-7. Semi-quantitative PCR analysis of DNA content of synchronized MEL cells.
Cells were isolated at each time point, DNA was extracted and PCR was
performed with primers to P-globin and necdin promoters. Equal amounts of
the two PCR products were run in the same lane and gels were stained with
SyBr Green. Bands were quantitated with Image Quant and the ratio of 3-
globin to necdin was calculated.
Recently extragenic transcripts were detected over the LCR and in between the
globin genes [156, 187, 202]. The generation of these transcripts was found to correlate
with stage-specific transcriptional activity and DNaseI sensitivity of P-globin locus
subdomains . This led to the hypothesis that transcription complexes are recruited
to various regions in the globin locus and track along the DNA in a unidirectional manner
thereby modifying chromatin structure. This is a plausible model because Pol II
transcription complexes are known to be associated with chromatin modifying activities
, and transcription through regulatory memory elements has been shown to be the
initial event in setting up activating or repressing functions of these elements in
Drosophila . Extragenic transcripts have also been detected in other complex gene
loci suggesting a broader role in the modulation of chromatin domains .
The role of the LCR in mediating high-level expression of the P-globin genes is
unquestionable. The LCR and the genes appear to be in close proximity in cells
expressing the globin genes, suggesting physical contacts between activities associated
with the LCR and proteins interacting with the promoter [120, 206]. How this contact
aides in stimulating transcription is not known, but could involve the transfer of activities
from the LCR to the globin genes . In the experiments described in this chapter we
analyzed the interaction of transcription complexes with LCR HS core elements and the
globin genes during S phase. The data demonstrate that RNA Pol II dissociates and re-
associates with the globin genes during replication, showing that the replication
machinery is capable of displacing transcription complexes. Interactions of NF-E2 and
USF appear to precede the re-association of Pol II with the globin locus suggesting that
these proteins assist in the recruitment of Pol II during S phase. The data thus support a
model in which the LCR and the globin genes cooperate in recruiting transcription
complexes to the globin locus.
We began our studies by addressing the question of whether RNA Pol II is
recruited to the LCR independent from its known interactions with the globin genes. The
ChIP assay per se does not differentiate between direct or indirect interactions. We show
here, that Pol II interacts with the HS2 core region together with other components of the
transcription initiation complex. This result shows that the crosslinking of Pol II to the
core is not solely due to elongating transcription complexes that are in the process of
generating the extragenic transcripts, but instead is due to Pol II that may be poised to
start extragenic transcription or ready for transfer to the gene promoters. This conclusion
is also supported by our observation that Pol II is not detectable in a region between the
HS2 and HS3 cores.
How does the polymerase interact with the LCR core HS sites? Is the
crosslinking to these sites due to interactions between HS2 and the globin genes, which
would be fixed in the ChIP assay, or is it due to direct association of transcription
complexes with the LCR HS core regions? We believe that the latter is true for the
following reasons: If the association of polymerase II with the HS2 core is due to protein
mediated interactions with the globin promoters, all proteins detected at the promoter
regions may also be detectable at the LCR and vice versa. This is clearly not the case.
For example, NF-E2 interacts strongly with HS2 in unsynchronized K562 cells, but only
weakly with the e-globin promoter. If the interaction of Pol II with HS2 is indirect and
caused by interactions with the e-globin promoter then antibodies against NF-E2 should
also efficiently precipitate the e-globin gene, as it is likely part of the same complex.
Although it is possible, but highly unlikely, that the mechanism of looping sequesters
proteins at either side of the interaction such that they are not detected on the other side
by formaldehyde crosslinking.
The results from the NRO analysis are consistent with the ChIP data and
demonstrate that transcription is ongoing at places where transcription complexes are
detected; supporting the notion that Pol II is recruited to the LCR independent from
interactions with the promoter regions. It is also interesting to note that while HS2 is
transcribed in both MEL and K562 cells, the HS2 5' flanking region in K562 cells is not,
consistent with the lack of Pol II binding to this region. This result suggests that
extragenic transcripts over the LCR do not represent long continuous transcripts but are
due to multiple transcription initiation sites in HS2, HS3, and possibly other regions of
the LCR. Our results are largely consistent with data recently published by Johnson et al.
 demonstrating that Pol II recruitment to the LCR is restricted to the core HS sites.
Sawado et al.  recently presented data demonstrating that an important
function of the mouse LCR is to stimulate globin gene expression at the level of
transcription elongation. The interpretation is based on the fact that in the absence of the
LCR no elongating transcription complexes were found in the transcribed region of the
pmaj-globin gene. If transcription complexes are stalled at the P-globin promoter in the
absence of the LCR, one might predict that more P-globin promoter bound Pol II is
detected in the mutant locus, compared with the wild-type locus. However, Sawado et al.
 found a two-fold reduction in promoter bound Pol II in the mutant allele suggesting
that another important function of the LCR is to mediate the recruitment of Pol II to the
P-globin genes. It is difficult to estimate the contribution of the LCR in recruiting Pol II
to the genes by comparing promoter-bound Pol II in promoters in which elongation takes
place with those in which elongation does not take place. It is thus possible that the
contribution of LCR mediated delivery of Pol II to the P-globin gene may have been
underestimated in these studies. Another possibility that should be considered is that in
the absence of the LCR unproductive transcription complexes are recruited to the globin
genes but that in the presence of the LCR these complexes are first recruited to the LCR,
rendered transcriptionally competent, and transferred to the globin genes .
The results from the cell synchronization/ChIP experiment described here are
consistent with a model proposing that an important function of the LCR is to represent a
reservoir of transcription and maybe other protein complexes, which may be used to
provide these activities to the globin genes during various stages of the cell cycle. NF-E2
may play a major role in directing these activities to the globin genes. In K562 cells NF-
E2 interacts directly or indirectly with HS2 and both e- and y-globin genes at specific
stages during S phase. Not surprisingly, the interaction of NF-E2 with the globin locus
precedes the re-association of Pol II, consistent with the notion that factors are needed to
recruit Pol II back to the locus after replication. We observed a similar pattern at the 3-
globin gene in MEL cells, although NF-E2 remains associated with HS2 and the P-globin
promoter, at least at the time points analyzed here. These data suggest an important role
of NF-E2 in recruiting Pol II to the globin locus, consistent with results from Johnson et
al.  demonstrating an impairment in Pol II/P-globin gene interactions in MEL cells
lacking NF-E2. In addition, it was recently shown that NF-E2 interacts with the human
P-globin locus control region before chromatin is re-modeled . The summary data
suggest that NF-E2 interacts with the globin locus at specific stages during the cell cycle,
prevents the formation of repressive chromatin structure, and facilitates the interaction of
transcription complexes with the globin locus.
Another interesting observation in these experiments is the interaction pattern of
USF2 during S phase in synchronized K562 cells. Previous studies have shown that USF
interacts with LCR core elements and with the P-globin promoter [143, 144, 163]. Our
data show that USF2 only interacts with HS2 and the globin genes at a stage at which
Pol II is not bound. It is conceivable that USF2, in conjunction with NF-E2, prepares
specific regions in the globin locus for the re-association of transcription complexes after
replication. Tolhuis et al.  recently proposed a model according to which all HS
sites in the globin locus participate in forming a specific chromatin conformation, called
the active chromatin hub (ACH). The interactions of NF-E2 and USF with the locus
control region and the genes at a stage in which Pol II is not bound in K562 cells (Figure
4-4, 2 hour time point) could indeed suggest that the LCR HS sites and the globin gene
promoters together orchestrate the recruitment of transcription complexes to the globin
gene locus, consistent with previous observations showing that enhancer and promoter
elements cooperate in the formation of DNaseI HS sites .
In summary, we have observed recruitment of Pol II directly to LCR HS2. We
also observed dynamic changes in the interaction of transcription factors and RNA-
polymerase II with the P-globin gene locus during S phase of the cell cycle. These
changes in interactions correlate with the occurrence of replication in these cells.
The results from the synchronization experiments are reproducible, with some
notable exceptions. In some experiments we detect NF-E2 at LCR HS2 at time 0 in
K562 cells. In addition, the recruitment of Pol II to the e-globin gene in K562 cells is
somewhat variable and we have performed experiments in which Pol II is crosslinked to
the e-globin gene after 6 hours. The variability may be due to subtle changes in the status
of the cells during synchronization.
ANALYSIS OF THE MECHANISM OF RNA POLYMERASE II TRANSFER FROM
THE LCR TO THE P-GLOBIN GENE PROMOTER
Erythroid-specific, high-level expression of the P-globin genes is regulated by the
locus control region (LCR), composed of multiple DNaseI hypersensitive (HS) sites and
located far upstream of the genes. Our lab and others have shown that the LCR core
elements recruit RNA Pol II [132, 169, 187-189]. The LCR HS sites are known to
interact with each other to form the LCR holocomplex. We proposed a model suggesting
that the LCR recruits Pol II and later transfers it to the gene promoters that are expressed
at specific developmental stages . Recent evidence obtained by the 3C technique
show that the LCR holocomplex interacts with whichever gene is transcriptionally active
at a particular stage in erythroid cells [117, 120, 158].
The 3C technology mentioned above is a method in which protein mediated DNA
interactions in live cells are captured by formaldehyde crosslinking. The crosslinked
chromatin is fragmented by restriction enzyme digestion and subsequently ligated under
conditions that favor intra-molecular ligation. These conditions allow DNA segments
that are normally located far away from each other to be ligated if they associate via
protein-protein and protein-DNA interactions. If these segments do not interact, the
probability of generating ligation products between these fragments is very low.
Formation of the ligation products is monitored by PCR.
Tolhuis et al. determined using 3C technology that the LCR HS sites interact with
each other, with the 3' HS site, and with the active globin genes in globin expressing cells
. The authors call the structure in which the LCR holocomplex interacts with an
active gene the active chromatin hub, while the intervening DNA loops out in agreement
with previously proposed looping models [101, 115]. The authors further showed that
during erythroid development in mice the 5', 3', and LCR hypersensitive sites interact
first to form a chromatin hub. The interaction of the genes with this structure then leads
to the formation of the active chromatin hub .
The data described above together with our own results from the synchronization
experiments support the LCR recruitment and transfer model. To test the Pol II transfer
hypothesis directly, we have developed a novel in vitro assay in which Pol II is first
recruited to an immobilized LCR construct and then transferred to a P-globin gene
promoter. We provide evidence that RNA polymerase can be transferred from the LCR
to a P-globin gene in vitro and demonstrate that the transfer is sensitive to mutations in
the P-globin gene promoter and facilitated by erythroid transcription factor NF-E2.
As a first attempt to examine the possibility that Pol II can be recruited to the LCR
and transferred to the P-globin gene, we established an in vitro system utilizing
immobilized LCR templates and nuclear extracts from IMEL cells (Figure 5-1).
A plasmid containing the entire human 3-globin LCR was linearized, biotinylated
and immobilized to streptavidin coated magnetic beads as described by Leach et al.
. After incubation with MEL nuclear extract the protein/DNA complex was washed
several times in binding buffer on a magnet. Then a plasmid containing the wild-type 3-
globin gene was added to the LCR and incubated for 10 minutes at 30 C in binding
buffer. The supernatant containing the P-globin gene was then removed from the
immobilized LCR, and subjected to crosslinking with formaldehyde and ChIP analysis
using a Pol II specific antibody as described in the Materials and Methods chapter (Figure
5-2 lane 2).
protein extract and
remove unbound proteins
ChIP and PCR
A schematic representation of the in vitro Pol II transfer experimental
procedure. The plasmid pLCR was linearized, biotin labeled, and
immobilized on streptavidin coated magnetic beads. After incubation with
MEL protein extracts, the LCR was placed on a magnet and all material not
interacting with the LCR was removed. The immobilized LCR/protein
complex was washed, resuspended in binding buffer and incubated in the
presence or absence of the P-globin gene. After placing the samples on the
magnet, the supernatant was removed and the sample containing the 3-
globin gene was crosslinked and subjected to ChIP analysis using an RNA
polymerase II specific antibody.
A second sample was processed in the same manner except that no Pol II specific
antibodies were added; this sample served as a negative control (Figure 5-2 lane 4). To
control for diffusion, we incubated 50 [tl binding buffer with the protein/LCR complex in
the absence of the P-globin template for 10 minutes at 30 C. The reaction was placed on
a magnet and the supernatant was removed and incubated with the P-globin template for
10 minutes at 30 C, after which formaldehyde was added to the reaction to induce
crosslinking. The sample was then subjected to ChIP analysis with Pol II antibodies
(Figure 5-2, lane 1).
+ 0 0
.1 h h.. I
to n '''t
I |in L"*
Q I- I- Z a.
1 2 3 4 5
Transfer of RNA polymerase II from immobilized LCR constructs to the 3-
globin gene. The sample containing the P-globin gene was crosslinked and
subjected to ChIP analysis using an RNA polymerase II specific antibody
(Lane 2). The supernatant of the sample incubated without the P-globin
gene for 10 minutes was later incubated with the P-globin gene template and
subjected to crosslinking and ChIP analysis using an RNA polymerase II
specific antibody (Lane 1). Lane 3 represents a reaction processed in the
same manner as described for lane 2 except that recombinant NF-E2 was
added to the reaction. Lane 4 represents a negative control in which no
antibody was added. Lane 5 represents the positive control in which the 3-
globin gene was incubated with MEL protein extract and subjected to ChIP
and PCR. Samples were processed as described in Fig 5-1 and subjected to
PCR analysis using a set of primers specific for the human 3-globin coding
region. The Pmaxi gene differs from the wild-type P-globin gene by 800 bp
and served as a PCR control in these experiments.
As a positive control, we incubated the P-globin gene for 30 minutes at 30 C with
MEL protein extract and subjected the sample to crosslinking and ChIP analysis (Figure
5-2, lane 5). After reversal of the crosslink and purification of the DNA, the samples
were analyzed by PCR using primers specific for the coding region of the P-globin gene.
The Pmaxi gene was added after the immunoprecipitation and served as a PCR control.
The result of this experiment demonstrates that Pol II can indeed be transferred
from the LCR to a P-globin template in trans (Figure 5-2, lane 2). The control
experiment shows that diffusion is negligible during the incubation time and does not
contribute significantly to the transfer of Pol II from the LCR to the P-globin gene
(Figure 5-2, lane 1). The inclusion of recombinant NF-E2 consistently enhanced the
transfer of Pol II from the LCR to the P-globin gene (Figure 5-2, lane 3) supporting
previous conclusions that NF-E2 is required for the recruitment of Pol II to the globin
In order to examine whether Pol II is specifically transferred to the P-globin gene
and not to sequences elsewhere in the plasmid we have performed the transfer
experiments in the presence of two restriction fragments derived from pRSPA/X, one
fragment containing the P-globin gene and the other containing the bacterial ampicillin
gene. The results are shown in Figure 5-3 and demonstrate that Pol II is specifically
transferred to the P-globin gene and not to other sequences present in the plasmid. In this
experiment we have increased the cycle number to visualize the signal corresponding to
the ampicillin gene and the P-globin gene in the no antibody control (lane 2). The data
show that the intensity of the P-globin signal increases in the presence of Pol II
antibodies, whereas the signal of the ampicillin gene remains unaltered.
To further analyze the specificity of the transfer, we performed the experiment with
several P-globin promoter mutants (described in ). In this experiment the Pmaxi gene
was added during the incubation with the protein/LCR complex and served as an internal
control. The results show that a mutant promoter in which the E-box element at +60 is
altered is impaired in recruiting Pol II (Figure 5-4).
1 2 3
Analysis of the specificity of Pol II transfer. The plasmid pRSPA/X was
digested with KpnI and SacI, incubated with the LCR/protein complex, and
processed as described in Figure 5-1 and 5-2. The precipitate was analyzed
by PCR using primers specific for the P-globin (3) or the bacterial
ampicillin gene (Amp), which are present on different restriction fragments.
Lane 1 shows a 100bp ladder.
Comparison of Pol II transfer using the wild-type P-globin gene (lane 1)
and a mutant P-globin gene (lane 2) in which an E-box at +60 was altered.
The samples were processed and analyzed as described in Fig 5-2 except
that the Pmaxi gene was added together with the wildtype or mutant 3-
globin gene and incubated with the immobilized LCR/protein complex.
The addition of BSA at the same concentration as NF-E2 did not enhance Pol II
transfer (Figure 5-5, lane 5), indicating that stimulation of Pol II transfer by NF-E2 is not
an unspecific effect due to increased protein content. To further examine whether Pol II
transfer is specific for the P-globin gene we have performed transfer experiments in the
presence of a P-globin gene lacking the TATA box. We have previously shown that this
mutant is inefficiently transcribed in vitro . We found that the P-globin TATA box
is necessary for Pol II recruitment (Figure 5-5, lane 6) demonstrating that Pol II is
specifically transferred to the P-globin gene and not to sequences elsewhere in the
0 + +
A-- P-m axi
w ., -g lob in
1 2 3 4 5 6
Figure 5-5. Transfer of RNA polymerase II to the P-globin gene in the presence of NF-
E2 or BSA. Samples were prepared and analyzed as described in panels
Figure 5-2 except that NF-E2, lane 4, or BSA, lane 5, were added during the
transfer step. Lane 6 represents transfer to a mutant P-globin gene template
lacking the TATA box.
To further test the specificity of the transfer we performed an experiment in which
we simultaneously incubated LCR/protein complexes with two different plasmids, one
containing LCR element HS2 and another containing the P-globin gene. The results in
Figure 5-6 show that Pol II is preferentially transferred to HS2 (lane 2), which could be
due to interactions between HS2 and the LCR/protein complex. Importantly, we
observed that in the presence of NF-E2, Pol II is preferentially transferred to the P-globin
gene and not to HS2 (lane 3). We also analyzed the transfer to the thymidine kinase
promoter and found that Pol II is not transferred to this template (compare Figure 5-6
lanes 1 and 4). We hypothesized that proteins in our erythroid cell extracts would bind
the thymidine kinase promoter since there are ubiquitously expressed proteins present in
these extracts that should bind other mammalian promoters such as the proteins of the
PIC. This again demonstrates that the transfer is specific for the P-globin gene. The
graph in Figure 5-7 summarizes the results of multiple experiments.
C C C C
0 Mi e 0 0 0
Z IMI U U
1 2 3 4 5 6 7
Analysis of NF-E2 mediated transfer to the P-globin gene. Samples were
processed as described in panel Figure 5-2 except that in lanes 2 and 3 the
transfer was carried out in the presence of both P-globin and HS2 containing
plasmids. In lane 3, NF-E2 was added to the transfer reaction. In lane 4,
transfer to the thymidine kinase/human growth hormone gene (TK-HGH)
construct was analyzed. Lane 1 represents the reaction of the no antibody
control containing all three plasmids (P-globin, HS2, and TK-HGH). Lanes
5, 6, and 7 represent positive PCR controls for 3-globin (lane 5), HS2 (lane
6) and TK-HGH (lane 7).
It was proposed that one function of the LCR could be to recruit macromolecular
protein complexes including Pol II, which are subsequently transferred to individual
globin genes in a developmental stage specific manner [96, 169, 179]. Similar
conclusions were derived from studies on the TCRP locus, which show that coactivators
and components of the transcription initiation complex are first recruited to an enhancer
element and subsequently delivered to the promoter .
We have developed an in vitro system to test the polymerase transfer hypothesis
and show that Pol II can indeed be transferred from an immobilized LCR template to the
P-globin gene (Figure 5-1). Moreover, the transfer of Pol II was enhanced by
recombinant NF-E2, supporting previous studies that implicate NF-E2 in the recruitment
of Pol II to the P-globin gene . The NF-E2 mediated stimulation of Pol II transfer is
not due to an increase in protein concentration as the same concentration ofBSA had no
5 Z -D
o + +
S I- I- iT
Q 20 z
X 0 K
Figure 5-7. Quantitative summary of the transfer experiments. The experiments were
repeated several times and the data were quantitated after gel electrophoresis
using SyBr green, Storm scanner, and Image Quant. The bars represent the
mean of the results of two to three experiments.
The observation that Pol II was not transferred to either the bacterial ampicillin
resistance gene or to the TK/HGH gene demonstrates that transfer is specific for the 3-
globin gene promoter (Figure 5-3 and Figure 5-6) when using extracts from erythroid
cells. It is likely that transfer would occur to other promoters as well if they contain
binding sites for proteins that mediate the transfer process, as I have shown to be the case
I also analyzed elements required for Pol II transfer to the P-globin promoter.
There was no transfer to the P-globin gene with a mutation at the TATA box and reduced
transfer to a promoter with a mutant +60 E-box (Figs. 5-4 and 5-5). This supports
previous data from our lab showing that the TATA box and +60 E-box are required for in
vitro transcription and suggest that these elements exert at least part of their function by
recruiting the polymerase .
The combined data suggest that the P-globin LCR, like other enhancer elements
, act in trans to assist in the recruitment of transcription complexes to the globin
genes. Although this in vitro system completely ignores nucleosome and higher order
chromatin structure, we believe that it does reveal some aspects of LCR function. It was
previously shown that specific characteristics of HS site formation and function,
including the recruitment of transcription complexes, can be recapitulated in vitro in the
absence of chromatin assembly . This demonstrates that chromatin is not directly
involved in recruiting transcription complexes to the globin locus but rather regulates a
preceding step, which is to render regulatory and functional sites available or unavailable
for the interaction with transcription factors and polymerase.
I show that Pol II can be transferred from immobilized LCR templates to globin
gene constructs in vitro and have established in vitro conditions for the analysis of RNA
polymerase transfer in the P-globin locus. In summary, we propose that transcription
complexes are first recruited to a highly accessible LCR holocomplex (Figure 5-8). The
genes are subsequently brought in contact with the LCR holocomplex, consistent with
data from conformational studies [120, 158], and transcription complexes are transferred
from the LCR to the promoter regions. This process is mediated by NF-E2 and possibly
other proteins (e.g. USF2). It is also facilitated by bringing strong globin gene promoters
(with high affinities for the transcription complex) in close proximity to LCR bound
s Gy Ay 6 p
Th RNA polymerase II
PV Preinitiation complex
Figure 5-8. Model of transcription complex recruitment to the P-globin gene. The LCR
and other HS sites have been shown to interact to form a chromatin hub. In
the active chromatin hub, the expressed genes interact with the HS sites. We
propose that transcription and other protein complexes are first recruited to a
highly accessible LCR holocomplex in the context of the proposed chromatin
hub. The genes come in close proximity to the LCR holocomplex by as yet
unknown mechanisms that may involve local remodeling of chromatin
structure at the active promoters. Transcriptions complexes are then
transferred from the LCR to high affinity binding sites at the globin gene
promoters. This transfer is facilitated by NF-E2 and/or related proteins.
DETERMINING THE FUNCTION OF THE USF, TFII-I AND HDAC3 PROTEINS IN
P-GLOBIN GENE REGULATION
At least two modes of regulation control the stage-specific expression of the globin
genes. The first parameter regulating stage-specific expression is the relative location of
the genes with respect to the LCR. When the order of the genes relative to the LCR is
reversed, the P-globin gene is inappropriately expressed at the embryonic stage .
The second parameter is the binding of stage-specific trans-acting factors to the
individual globin promoter regions leading to activation of transcription at the appropriate
times during development. A well-characterized example for this mode of regulation is
represented by the transcription factor EKLF, which specifically activates expression of
the adult P-globin gene during definitive erythropoiesis [137-139, 142]. EKLF exerts part
of its activating function by recruiting chromatin-remodeling complexes to the P-globin
gene promoter, which leads in turn to a local change in nucleosome organization .
The basal promoter of the human adult P-globin gene is composed of a TATA-like
element and an initiator sequence, both of which were shown to interact with the TFII-D
protein complex [211, 212]. In addition to components of the TFIID complex
(specifically, TAFII250 and TAFII150), a diverse group of other proteins can also
interact with initiator sequences [150, 170]. Notable among these are the helix-loop-
helix proteins USF and TFII-I, because both proteins have been implicated in the
recruitment of transcription complexes to TATA-less promoters and in the stabilization of
transcription complexes in TATA-box-containing promoters [145, 146, 213].
Regulatory DNA elements located downstream of transcription initiation sites
(downstream promoter elements, DPE) that contribute significantly to the formation of
stable transcription complexes have been well characterized in a variety of Drosophila
genes . In some cases these downstream promoter elements function in the absence
of TATA elements and cooperate with initiator sequences to recruit transcription
complexes. The transcription cofactor NC2 (negative cofactor 2), originally identified as
a repressor of RNA Pol II transcription, activates transcription of DPE-containing
promoters and inhibits transcription of TATA-box-containing promoters in vitro .
The human 3-globin gene contains two conserved E-box motifs located 3' to the
initiator sequence. The first E-box overlaps the initiator while the second E-box is
located 60 bp 3' to the transcription initiation site (Figure 6-1). Previous results in the lab
demonstrate that the human P-globin promoter's TATA-like motif, initiator and
downstream E-box element are all required for high level transcription in vitro .
This indicates that E-box motifs contribute to the efficient formation of transcription
complexes on the adult P-globin gene. The same study also indicated that USF 1 and
TFII-I interact with the initiator and overlapping E-box, while USF 1 and USF2 interact
with the downstream E-box in vitro .
Here I investigated the interactions of the HLH proteins in vivo. The data show
that USF 1, USF2, TFII-I and p45 interact with the P-globin promoter in vivo, supporting
the previous in vitro results . By comparing the protein-DNA interactions on
expressed and non-expressed P-globin gene promoters, it was discovered that USF2 and
TFII-I interact more efficiently with the non-expressed than with the expressed gene.
The interactions of USF 1 and p45 with the P-globin promoter are restricted to erythroid
cells expressing the P-globin gene. Further, TFII-I was implicated in recruiting HDAC3
to the human P-globin gene promoter in cells that are not expressing the gene. The
results indicate that helix-loop-helix proteins contribute to the formation of transcription
complexes on the adult human 3-globin gene, and suggest that the differential association
of HLH proteins with the basal promoter contribute to the stage-specific expression of the
INII Mi *2 FI -bs +60 E-bra
BE: ttGCijt tc ttcq*ev u cl iq3ttticactaLa DDtcaacag- -Ciaao ct L
H:ctcCt Ua Ltt gLat Lgac&t tngLltu taatiact n.cgaaCaSr -gt-cA tcatqi;trcct
Pr-ctqckt tttMoc-fg ^tccalcjaact tttactgZC* tcccuca cgai'm cgq t: cncta
Figure 6-1. Sequence alignment of the human 3-globin downstream promoter region.
Shown are three sequences of the adult P-globin downstream promoter
region from human (H), mouse (M) and rabbit (R). Shaded boxes highlight
the position of E-box motifs (CANNTG). Two of these E-boxes, the one
overlapping the initiator and the distal E-box, are conserved in all three
species, whereas the E-box located at +20 is only present in the human and
rabbit genes. The open box delineates the position of the MARE-like
In these studies I used two erythroid cell lines, MEL, a murine erythroleukemia cell
line expressing the adult P-globin gene, and K562, a human erythroleukemia cell line
expressing the embryonic 3-type globin genes. I examined whether USF 1, USF2 or
TFII-I interact with the P-globin gene in vivo by performing ChIP experiments in MEL
and in K562 cells. The results of these experiments are shown in Figure 6-2 and
demonstrate that both USF1 and USF2 could be crosslinked to the P-globin gene
promoter in MEL cells in vivo (Figure 6-2, lanes 4 and 5). TFII-I could also be
crosslinked to the globin promoter (Figure 6-2, lane 6); however, the signal for TFII-I
was consistently weaker in MEL versus K562 cells. Antibodies recognizing the p45
subunit of NF-E2 and acetylated histone H3 also precipitated the P-globin gene in MEL
cells (Figure 6-2, lanes 7 and 8, respectively). The precipitation with acetylated H3
antibodies was expected from previous studies .
To analyze whether the interaction ofUSF proteins with the P-globin promoter
could be correlated with the transcriptional activity of the gene I carried out ChIP
analyses in K562 cells, in which the P-globin gene is not expressed. In contrast to our
observations in MEL cells, USF 1 does not interact with the P-globin gene in K562 cells,
whereas TFII-I and USF2 do interact (Figure 6-2, lanes 5 and 6). The interactions of p45
(NF-E2) and acetylated H3 with the P-globin gene were also less efficient in K562 cells
when compared to MEL cells (compare Figure 6-2, lanes 7 and 8, MEL versus K562). In
control experiments I amplified a region within the LCR located upstream ofHS2 (HS2 5'
flank; Figure 6-2, lanes 10-16). The results demonstrate that the observed interactions of
USF1, USF2, TFII-I and p45 in erythroid cells are specific for the P-globin gene.
To analyze the possibility that the differential binding ofUSF1, NF-E2 and, to a
lesser extent, TFII-I is due to differential expression of these proteins in the two cell lines,
Kelly Leach in our laboratory carried out western blotting experiments using antibodies
specific for USF1, USF2, TFII-I and NF-E2. The results show that USF1 and USF2 are
expressed at similar levels in K562 and MEL cells. However, TFII-I is expressed at
higher levels in K562 cells and p45 (NF-E2) is expressed at higher levels in MEL cells.
P-globin HS2 5'flank
MEL o S
1 2 3 4 5 6 7 8 9 10111213141516
Figure 6-2. Characterization of protein-DNA interactions in the human P-globin
downstream promoter region in vivo. Chromatin immunoprecipitation
(ChIP) experiment analyzing the interaction of proteins with the
murine/human P-globin gene or the HS2 5'flanking region in MEL and
K562 cells in vivo. Complexes were precipitated with either no antibody
(lanes 3 and 11) or with antibodies specific for USF1 (lanes 4 and 12), USF2
(lanes 5 and 13), TFII-I (lanes 6 and 14), NF-E2 (p45, lanes 7 and 15) or
acetylated histone H3 (lanes 8 and 16). DNA was purified from the
precipitate and analyzed by PCR for the presence of the murine or human P-
globin gene (210 and 321 bp, respectively, lanes 2-8) or the murine or
human HS2 5'flank (336 and 565 bp, respectively, lanes 10-16). As positive
controls, the input DNA was also analyzed by PCR (lanes 2 and 10). Lanes
1 and 9 show 100 bp markers.
Based on previously published data , I investigated potential interactions
between TFII-I and HDAC3 with each other and the promoter of the silenced P-globin
gene in K562 cells. For this, I carried out chromatin double immunoprecipitation
(ChDIP) experiments, in which two subsequent ChIP reactions with two different
antibodies are performed. In the ChDIP, the first ChIP assay was performed with an
antibody against TFII-I. The protein-DNA complexes eluted from the beads were then
used as substrate for another CHIP assay using an antibody against HDAC3. The results
show that in the first round of immunoprecipitation TFII-I was bound to the P-globin
gene promoter at levels above that of the no antibody background (Figure 6-3). In the
subsequent round of immunoprecipitation, HDAC3 was detected in P-globin templates
bound by TFII-I.
In contrast there were no e-globin promoter templates detected in both TFII-I and
HDAC3 precipitates. As a negative control, I also analyzed the linker region between
HS2 and HS3 and conclude that TFII-I and HDAC3 do not interact with this region. The
no antibody background after the second round of immunoprecipitation was drastically
decreased compared to the first one, probably due to the loss of non-specific interactions
with the beads.
P gene a gene HS2&3 Linker
IS I 4
Figure 6-3. Chromatin double immunoprecipitation (ChDIP) experiment analyzing the
interaction of TFII-I and HDAC3 at different regions within the human 3-
globin locus. Complexes were first immunoprecipitated with an antibody
against TFII-I and then these TFII-I selected complexes were
immunoprecipitated with an antibody against HDAC3.
Although the ChDIP experiment showed that TFII-I and HDAC3 are both bound to
the promoter of the non-expressed 3-globin gene, it did not provide evidence of direct
interactions between these proteins. In order to examine the possibility that TFII-I
directly interacts with HDAC3, we carried out co-immunoprecipitation experiments.
Whole cell extracts from K562 cells were immunoprecipitated with antibodies against
USF 1, TFII-I, Pol II, USF2, HDAC3 and a no antibody negative control. The complexes
eluted off the beads were separated by SDS-PAGE and transferred to a nylon membrane
for Western blot analysis. An antibody against HDAC3 was used for the Western
Supporting our previous observation, we detected HDAC3 only in complexes first
immunoprecipitated with antibodies against TFII-I (Figure 6-4). In contrast, HDAC3
was not detected in the samples containing USF USF2 or Pol II. This shows that TFII-I
and HDAC3 directly interact in K562 cells.
1 ^ ..4 HDAC3
Figure 6-4. Co-immunoprecipitation experiment analyzing the interaction of TFII-I and
HDAC3. Whole cell extracts from K562 cells were first
immunoprecipitated with the indicated antibodies then the complexes pulled
down were run on SDS-PAGE and probed by with HDAC3 antibodies in a
The promoter of the human P-globin gene is characterized by the presence of a non-
canonical TATA-box (CATAAA) located 25-30 bp upstream of the transcription start
site . Deviation from consensus TATA sequences often weakens promoters and
leads to the requirement of additional elements for the efficient recruitment or
stabilization of transcription complexes. For example, a recent report demonstrated that
the TBP/TFII-A pre-initiation complex interacts only weakly with the CATAA motif
. An additional sequence that contributes to high level P-globin gene transcription is
an initiator element located at the transcription start site. Lewis et al. [211, 212] have
shown that this element interacts with the TFII-D complex in vitro and that mutations in
this sequence reduce transcription efficiency.
In addition to the initiator element, the P-globin gene contains several E-box motifs
downstream of the transcription start site  (Figure 6-1). Two of these elements are
conserved across species. Because previous studies have shown that E-box-binding
proteins can participate in the formation of transcription complexes on both TATA-
containing and TATA-less genes [145, 146, 213], we were interested in examining the
functional role of these proteins in P-globin gene transcription.
The P-globin initiator/E-box interacts with TFII-Iand USF1, whereas the distal E-
box at +60 interacts with USF1 and USF2 in vitro . There are a number of reports
in the literature documenting the interaction of USF and TFII-I with sequence elements in
the vicinity, mostly downstream, of transcription initiation sites [219-222]. Hsieh et al.
 recently identified an E-box motif located downstream of the APEG-1 gene from
+39 to +44. This element interacts with USF proteins and mutation of this E-box
dramatically reduces transcription.
TFII-I interacts with the P-globin initiator in vitro and in vivo and may play a role
in activating transcription of the P-globin gene in MEL cells. However, several
observations suggest that TFII-I is not required for P-globin gene transcription. First,
antibodies against TFII-I were not able to inhibit transcription of the P-globin gene in
vitro . Secondly, crosslinking of TFII-I to the P-globin gene in adult erythroid cells
is inefficient compared to USF 1 and USF2 (Figure 6-2). The ChIP assay does not provide
information as to what fraction of cells have a specific protein bound at a specific site. It
is thus possible that the weak PCR signal from TFII-I-selected MEL chromatin fragments
indicates that only a small fraction of cells have TFII-I bound at the initiator. This
fraction of cells could represent non-expressing cells. This is consistent with our
observation that TFII-I crosslinks more efficiently to the P-globin gene in K562 cells
(Figure 6-2A), suggesting that TFII-I may only be able to interact with the P-globin gene
in the absence of transcription complex formation.
We have performed RT-PCR (not shown) and western blotting experiments to
characterize the expression pattern of TFII-I in K562 versus VMEL cells. The RT-PCR
analysis revealed that TFII-I is transcribed with equal efficiency in both cell types.
However, western blots of nuclear extracts consistently showed that full-length TFII-I
protein is present at higher levels in K562 than it is in VMEL nuclear extracts. The reason
for the lower amount of TFII-I in MEL nuclear extracts may be due to post-
transcriptional mechanisms. This lower level of TFII-I in VMEL cells may contribute to
the lower level of TFII-I bound to the P-globin gene promoter in these cells.
Whether the binding of TFII-I to the P-globin promoter in erythroid cells with
embryonic phenotype contributes to repression of the P-globin gene or simply reflects the
absence of TFII-D binding was an issue that we sought to resolve. A recent report shows
that TFII-I interacts with histone deacetylase 3 . It is thus possible that TFII-I
interacts with the P-globin gene in embryonic erythroid cells, recruits histone
deacetylases and contributes to the repression of P-globin gene expression by rendering
the chromatin structure inaccessible. Our ChIP data clearly show a strong reduction of
acetylated H3 binding to the P-globin promoter in K562 cells compared to MEL cells
(Figure 6-2, lane 8).
The combination of ChDIP and co-immunoprecipitation experiments indicated that
TFII-I and HDAC3 interact with each other and associate with the P-globin gene in non-
expressing erythroid cells. The data support a model according to which TFII-I recruits
HDAC3 to alter the chromatin structure and therefore to silence the P-globin gene
Our data led us to formulate a hypothesis on the function of proteins interacting
with the P-globin downstream promoter during development (Figure 6-5). We propose
that TFII-D, NF-E2, USF 1 and USF2 interact with the globin promoter in adult erythroid
cells and provide a platform for the efficient recruitment of transcription complexes [179,
195, 211, 212]. The functional role of NF-E2 in this process remains to be established,
but other data presented before (Chapter 5) suggest that it may be involved in the transfer
of RNA Pol II from the LCR to the gene promoter.
The interaction of TFII-I and USF2 with the P-globin promoter in K562 cells could
contribute to the repression of this gene at the embryonic stage of erythroid development
by recruiting histone deacetylase activities. Other studies also implicate USF2 as a
potential transcriptional repressor . However, in adult cells, USF2 and USF1
interact with the E-box at +60 and stimulate transcription. This conclusion is supported
by our previous observation that in vitro transcription of the P-globin gene is inhibited by
pre-incubating the protein extracts with USF2 antibodies .
In summary, our data provide evidence that HLH proteins differentially interact
with the initiator and downstream promoter during erythroid development. The
combination of USF 1, USF2, NF-E2 and, perhaps, TFII-I contribute to expression of the
P-globin gene. Conversely, interactions of TFII-I along with USF2 recruit HDAC3
within the P-globin gene, leading to the repression of the P-globin gene.
TATA Initiator MARE +60Ebox
HDAC3 TFII I/USF2
Figure 6-5. Summary and model of protein-DNA interactions at the human P-globin
promoter. The human 3-globin promoter consists of a TATA-box and an
initiator. These sequences are bound by the TFII-D complex in cells
expressing the P-globin gene (active). Additional sequences required for the
formation of active transcription complexes are MARE and E-box elements
located in the downstream promoter region. These sequences are bound by
NF-E2 (p45 and p18) and USF, respectively. It is proposed that in cells not
expressing the P-globin gene (inactive), TFII-D is not bound and the
initiator sequence is occupied by protein complexes consisting of TFII-I and
CONCLUSIONS AND FUTURE DIRECTIONS
We have established a novel technique that allows the identification of sequences to
which a specific protein binds in the context of chromatin. This represents a powerful
tool for discovering binding sites of proteins not only in the P-globin locus but also in
other areas of the genome. We used this technique to demonstrate that NF-E2 binds to a
non-consensus MARE in the mouse P-globin downstream promoter region.
This technique could be refined by optimizing the restriction enzyme digestion
step. I noticed the appearance of unspecific DNA degradation products after long
incubation times with the restriction enzymes. The Grosveld group followed a similar
digestion protocol for their 3C experiments except that they utilized SDS to remove any
non-crosslinked proteins from the DNA and Triton X-100 to sequester SDS for more
efficient digestion . However, whether this is applicable for the ChIP footprint
technique is questionable because the following step after fragmentation is
immunoprecipitation and not, as is the case in the 3C method, DNA ligation.
The ChIP footprinting technique could be used to further identify binding sites for
several of the proteins our lab investigates. For example, we could analyze whether TFII-
I binds to the P-globin initiator or to the downstream E-box elements in vivo. Similarly
we could investigate a possible difference in USF binding between 3-globin promoters of
expressing and non-expressing cells.
ChIP within the P-Globin Locus
Using the ChIP assay, I showed that RNA Pol II is directly recruited to HS2 of the
LCR. As expected, Pol II and associated transcription factors are also present at the
promoters of expressed genes in MEL and K562 cells. Acetylated histone H3 was also
present at the expressed genes, as has been previously documented [103, 155]. NF-E2
was consistently seen to associate with the promoters of the active genes in the locus.
Having optimized the ChIP assay for use within the mouse and human P-globin locus of
MEL and K562 cells respectively, there are many more experiments that can be carried
out utilizing this assay.
Another group using a MEL cell line that needs DMSO induction for high level 3-
globin expression has shown that NF-E2 only binds to the LCR and gene promoter after
DMSO induction . It appears that our MEL cells are quite different since DMSO is
not required for 3-globin gene expression or NF-E2 binding as shown in Figure 4-3. It
would be interesting to repeat the ChIP experiment shown in Figure 4-3 with DMSO
induced MEL cells to see if there are any differences in binding between induced and
There is currently data available on the binding of the factors Bachl, GATA1,
GATA2, CREB binding protein (CBP), EKLF, Spl and YY1 to the P-globin locus. It
may be possible to use the ChIP assay in association with real-time quantitative PCR to
elucidate differences in binding of these factors between the transcribed and non-
transcribed regions of the locus in either MEL and K562 cells that has not been