Identification and Characterization of Novel Sir3/MeCP2 ...
Transcript of Identification and Characterization of Novel Sir3/MeCP2 ...
Marshall UniversityMarshall Digital Scholar
Theses, Dissertations and Capstones
1-1-2009
Identification and Characterization of Novel Sir3/MeCP2-Chromatin InteractionsNicholas L. Adkins
Follow this and additional works at: http://mds.marshall.edu/etdPart of the Biochemical Phenomena, Metabolism, and Nutrition Commons, Biological
Phenomena, Cell Phenomena, and Immunity Commons, Genetic Phenomena Commons, GeneticProcesses Commons, Genetic Structures Commons, and the Medical Genetics Commons
This Dissertation is brought to you for free and open access by Marshall Digital Scholar. It has been accepted for inclusion in Theses, Dissertations andCapstones by an authorized administrator of Marshall Digital Scholar. For more information, please contact [email protected].
Recommended CitationAdkins, Nicholas L., "Identification and Characterization of Novel Sir3/MeCP2-Chromatin Interactions" (2009). Theses, Dissertationsand Capstones. Paper 392.
1
Identification and Characterization of Novel Sir3/MeCP2-Chromatin Interactions
Dissertation submitted to the
Graduate College of
Marshall University
in partial fulfillment of the requirements for
the degree of
Doctor of Philosophy
in
Biomedical Sciences
by
Nicholas L. Adkins
Approved by
Dr. Philippe Georgel, Ph.D., Committee Chairperson
Dr. Eric Blough, Ph.D.
Dr. Elaine Hardman, Ph.D.
Dr. Richard Niles, Ph.D.
Dr. Vincent Sollars, Ph.D.
Department of Biochemistry, Marshall University School of Medicine
October 2009
ii
ABSTRACT
Identification and Characterization of Novel Sir3/MeCP2-Chromatin Interactions
By Nicholas L. Adkins
The eukaryotic genome is packaged into chromosomes that are made up of a highly
organized and heavily regulated structure called chromatin. The proteins involved in the
compaction of DNA into this condensed state are mostly understood at the level of the structure
of the nucleosome. The higher order arrangement of chromatin and how it effects gene
regulation is only partially understood and characterized. The compaction of nucleosomal arrays
into 30-nm and higher structures are partially the responsibility of architectural, or structural,
chromatin associated proteins. The following dissertation analyzes the individual chromatin
contributions of two well studied architectural proteins, the yeast silencing protein Silent
Information Regulator 3 (Sir3) and the human transcriptional regulator methyl CpG binding
protein 2 (MeCP2). Silencing in yeast is the responsibility of the SIR family of proteins.
Classically, the Sir3 protein has been characterized as associating with chromatin through the
hypo-acetylated N-termini of the core histones H3 and H4. The Sir3 protein has recently been
found to contain a DNA-binding element, my studies characterized Sir3-nucleic acid interactions
and showed that Sir3 can bind to chromatin independently of histone N-termini. In contrast, the
MeCP2 protein has classically been characterized as a methylated DNA dependent
transcriptional repressor, but recent genome-wide analysis reveals MeCP2 distribution can occur
on promoters of active genes. Recent in vitro work with MeCP2 and nucleosomal arrays showed
a highly ordered, compacted chromatin structure even in the absence of DNA methylation.
MeCP2 is of particular biological interest due to the observed link with the neurodevelopmental
iii
disorder Rett Syndrome (RTT). My studies demonstrated that MeCP2 can bind in vitro to the N-
termini of core histones H2A, H3, and H4. Additionally, the removal of these tails impacted
MeCP2-chromatin interactions, and resulted in a reduced level of nucleosomal array
condensation. Importantly, the two RTT mutants analyzed here, R133C and R168X, exhibited
differential binding to histone N-termini. These results add to the understanding of chromatin
organization and arrangement by demonstrating and characterizing additional chromatin contacts
for these two chromatin associated proteins.
iv
ACKNOWLEDGMENTS
I want to express my gratitude to all of my wonderful science teachers and advisors
throughout my life. First and foremost, I want to thank my PhD advisor, Dr. Philippe Georgel,
for his guidance, patience, and help in both my professional and personal life. Without his
influence I would not be where I am at today. I also want to thank all of my committee members
for their outstanding guidance during my graduate studies. I also want to express gratitude to Dr.
Elizabeth Murray and Dr. Mike Little for all of the help during my undergraduate years.
I especially want to thank all of my high school and junior high school teachers without
whom I would not have had any interest of going into my current profession. Mr. Walker, my
junior high school science teacher at Rupert, WV, will always be remembered for his penchant
for bringing wit and humor to the classroom. Mr. Lewis and Mr. Meadows were great high
school biology teachers at Greenbrier West High School. I am incredibly lucky to have had both
in my school system at the same time. And I want to thank Mr. Arbuckle, an outstanding
chemistry and physics teacher at Greenbrier East High School, whose kindness to the new kid in
school will never be forgotten.
Thank you all for your advice, patience, and belief in me.
v
LIST OF FIGURES
Figure 1.1: X-ray Crystal Structure of the Nucleosome Core Particle (NCP) .............................................. 6 Figure 1.2: Models of the 30-nm Chromatin Fiber ....................................................................................... 8 Figure 1.3: Different Mechanisms of Epigenetic Regulation ..................................................................... 10 Figure 1.4: Mechanism of DNA Methylation ............................................................................................. 11 Figure 1.5: DNA Methylation ..................................................................................................................... 12 Figure 1.6: Sir3 Domain and Partner Interaction Location ......................................................................... 29 Figure 1.7: The Known Members of the Mammalian methyl-CpG-binding Protein Family ..................... 33 Figure 1.8: The Described Domains of the Methyl-CpG-binding Protein 2 (MeCP2) ............................... 42 Figure 1.9: Location and Frequency of MeCP2 “Hotspot” Mutations in RTT Patients ............................. 46 Figure 2.1: Sir3p binding efficiency to DNA is similar to that of tailless NA ............................................ 59 Figure 2.2, A: Binding of Sir3p as a function of DNA length. ................................................................... 60 Figure 2.2, B: Binding of Sir3p to short DNA fragments. .......................................................................... 62 Figure 2.3, A and B: Predicted curvature and EMSA analysis of Sir3p binding to various DNA conformations ............................................................................................................................................. 65 Figure 2.3, C: Fraction of free DNA for increasing concentrations of Sir3 protein ................................... 68 Figure 2.4: Sir3p binds cooperatively to DNA. .......................................................................................... 69 Figure 2.5: EM imaging of Sir3p-DNA complexes at Low and High DNA concentrations. ..................... 71 Figure 2.6: Effect of DNA concentration on Sir3p binding. ....................................................................... 73 Figure 2.7: Sir3p binds to ssDNA. .............................................................................................................. 74 Figure 2.8: Model for DNA-mediated Sir3p binding component. .............................................................. 78 Figure 3.1: MeCP2 Domains and RTT Mutants. ........................................................................................ 87 Figure 3.2, A and B: MeCP2 binding is influenced by histone N-termini. ................................................. 92 Figure 3.2, C: EM imaging of MeCP2-NA complexes with and without histone N-termini tails. ............. 93 Figure 3.3: MeCP2 binding efficiency to composite NAs lacking one type of histone N-termini. ............ 95 Figure 3.4: MeCP2 interacts with the N-termini of H2A, H3, and H4 in a GST-fusion pull-down assay. 97 Figure 3.5, A: Effects of histone acetylation on MeCP2 binding. .............................................................. 98 Figure 3.5, B: Effects of histone acetylation on MeCP2 compaction. ........................................................ 99 Figure 3.6, A: RTT-linked MeCP2 R133C differential chromatin interaction. ........................................ 101 Figure 3.6, B: GST-fusion pull-down with MeCP2 R133C. ..................................................................... 102 Figure 3.7: The RTT-linked truncated MeCP2 R168X mutant loses histone N-termini interaction. ....... 103 Figure 4.1: Sir3’s Protein Primary Amino Acid Sequence. ...................................................................... 122 Figure 4.2: Degree of Intrinsic Disorder in Sir3’s Primary Structure. ...................................................... 123 Figure 4.3: MeCP2’s Protein Primary Amino Acid Sequence. ................................................................ 125 Figure 4.4: Degree of Intrinsic Disorder in MeCP2’s Primary Structure. ................................................ 125 Figure 4-5: Alignment between Sir3’s known Domains and Calculated Intrinsic Disorder Regions. ..... 127 Figure 4.6: Alignment between Sir3’s known Domains and Calculated Intrinsic Disorder Regions. ...... 129
vi
LIST OF TABLES
Table 1.1: Post-Translational Modifications Effects on Transcription ....................................................... 14 Table 1.2: Function and Location of Variant Histones ............................................................................... 20 Table 2.1: Characterization of mobility and Sir3 binding of the DNA fragments ...................................... 70
vii
LIST OF ABBREVIATIONS
aa amino acid Arf ADP Ribosylation Factor BAH bromo-adjacent homology bp base pairs BRCA1 breast cancer 1 gene CAP chromatin associated protein CD circular dichroism CHD chromodomain ChIP-chip chromatin immunoprecipitation-microarray chip technology CREB1 cAMP responsible element binding protein 1 CSD chromoshadow domain DNMT DNA methyltransferases dsDNA double-stranded DNA EMSA electrophoretic mobility shift assay EM Electron Microscopy FRAP fluorescence recovery after photobleaching FWJ four-way-junction GFP green fluorescent protein HAT histone acetyltransferase HDAC histone deacetylases Hdm2 human double minute 2 HEGN 10 mM HEPES, 0.1 mM EDTA, 5% glycerol, 10 mM KCL, 0.1% NP-40
viii
HMG high mobility group HP1 heterochromatin protein 1 H3K9me methylation of histone H3 lysine 9 H3K9 H3 lysine 9 H3S10 phosphorylation on histone H3 serine 10 H5EGN 5 mM HEPES, 0.1 mM EDTA, 5% glycerol, 10 mM KCL, 0.1% NP-40 ID intrinsically disordered MBD methylated DNA binding proteins MBP methyl-DNA binding protein MeCP1 methyl-CpG-binding protein 1 MeCP2 methyl-DNA binding protein 2 MENT Myeloid and Erythroid Nuclear Termination stage-specific protein NA nucleosomal arrays NBD nucleosomal binding domain NCP Nucleosome Core Particle N-CoR nuclear receptor co-repressor NuRD nucleosome remodeling and deacetylation PcG polycomb group protein complex PEV position effect variegation POZ/BTB poxvirus and zinc finger/Bric-a-brac, Tramtrack, Broad-complex PTM post-translational modification QAGE Quantitative Agarose Gel Electrophoresis Rap1 repressor activator protein 1
ix
rDNA ribosomal DNA RTT Rett syndrome R106W Arginine to Tryptophan Point Mutation at Residue 106 R133C Arginine to Cysteine Point Mutation at Residue 133 R168X Arginine to Stop Codon at Residue 168 R294X Arginine to Stop Codon at Residue 294 R255X Arginine to Stop Codon at Residue 255 R306C Arginine to Cysteine Point Mutation at Residue 306 SAM S-adenosylmethionine SHL super-helical Sir silent information regulator Sir3 silencing information regulator 3 TPE telomere position effect TRD transcription repression domain T158M Threonine to Methionine Point mutation at Residue 158 YprA yeast proteinase A wt wild type
x
TABLE OF CONTENTS
ABSTRACT ...................................................................................................................................................... ii
ACKNOWLEDGMENTS .................................................................................................................................. iv
LIST OF FIGURES ............................................................................................................................................ v
LIST OF TABLES ............................................................................................................................................. vi
LIST OF ABBREVIATIONS ............................................................................................................................. vii
TABLE OF CONTENTS ..................................................................................................................................... x
CHAPTER ONE: BACKGROUND AND SIGNIFIGANCE ..................................................................................... 1
INTRODUCTION ......................................................................................................................................... 1
CHROMATIN COMPOSITION AND ORGANIZATION ..................................................................... 3
EPIGENETIC MODIFICATIONS ........................................................................................................... 9
DNA-METHYLATION ......................................................................................................................... 10
HISTONE POST-TRANSLATIONAL MODIFICATIONS .................................................................. 13
HISTONE VARIANTS .......................................................................................................................... 18
CHROMATIN-ASSOCIATED PROTEINS .......................................................................................... 20
EUKARYOTIC GENE SILENCING ..................................................................................................... 26
SIR FAMILY OF PROTEINS ................................................................................................................ 27
SILENCING INFORMATION REGULATOR 3 (SIR3) ....................................................................... 28
POST-TRANSLATIONAL MODIFICATIONS OF SIR3 .................................................................... 32
METHYLATED-DNA BINDING PROTEIN FAMILY ....................................................................... 32
METHYL-CpG-BINDING PROTEIN 2 (MeCP2) ................................................................................ 35
MeCP2 PATHOLOGY (RTT AND CANCER) ..................................................................................... 36
ISOFORMS OF MeCP2 ......................................................................................................................... 38
FUNCTION AND STRUCTURE OF MeCP2 ....................................................................................... 38
POST-TRANSLATIONAL MODIFICATION OF MeCP2 ................................................................... 43
MeCP2 RTT MUTATIONS ................................................................................................................... 44
CHAPTER 2: ROLE OF NUCLEIC ACID BINDING IN SIR3p‐DEPENDENT INTERACTIONS WITH CHROMATIN
FIBERS ......................................................................................................................................................... 47
ABSTRACT ............................................................................................................................................ 48
INTRODUCTION .................................................................................................................................. 49
xi
EXPERIMENTAL PROCEDURES ....................................................................................................... 51
RESULTS ............................................................................................................................................... 57
DISCUSSION ......................................................................................................................................... 75
CHAPTER 3: DIFFERENTIAL BINDING OF WILD TYPE AND RTT MeCP2 MUTANTS TO HISTONE TAILS
MEDIATES CHROMATIN INTERACTIONS. .................................................................................................... 82
ABSTRACT ............................................................................................................................................ 83
INTRODUCTION .................................................................................................................................. 84
EXPERIMENTAL PROCEDURES ....................................................................................................... 88
RESULTS ............................................................................................................................................... 90
DISCUSSION ....................................................................................................................................... 103
CHAPTER 4: DISCUSSION, CONCLUSION, AND FUTURE STUDIES ............................................................. 112
DISCUSSION ....................................................................................................................................... 112
CONCLUSIONS ................................................................................................................................... 126
FUTURE STUDIES .............................................................................................................................. 130
REFERENCES .............................................................................................................................................. 135
1
CHAPTER ONE: BACKGROUND AND SIGNIFIGANCE
INTRODUCTION
Recent advances in research on the mechanisms of gene expression and silencing in
eukaryotic organisms have revealed an incredible level of complexity. In addition to regulatory
DNA sequences, chromatin structure has been shown to be intimately linked to the control of
gene expression with coordinated rearrangement of multiple molecules being required for DNA
accessibility. In contrast to the general dogma that an open, accessible chromatin structure
facilitates gene activation, silencing requires the compaction of chromatin over areas of the
genome where genes are repressed. These transitions from unfolded to compacted chromatin
must also occur in a reversible manner, suggesting coordinated regulatory events. Unintended
alterations in structure can often lead to disease through anomalous gene expression. Aberrant
changes in structure can occur through several different mechanisms including altered patterns of
DNA-methylation, improper post-translational modifications (PTMs) of DNA-associated
proteins such as histones, or genetic mutations that result in changes for regulatory factors
binding affinity. Treatment of diseases linked to these changes is critically dependent on
understanding all of the protein-protein or DNA-protein interactions that influence chromatin
dynamics and/or any signals that result in structural changes. The study of chromatin associated
proteins (CAPs) and their interactions with chromatin is therefore essential, and is one of the
fundamental first steps towards understanding the pathological mechanisms of certain diseases.
CAPs’ functions are regulated by similar protein-protein interaction domains, catalytic
subunits, and DNA-binding motifs. In addition, CAPs share similar characteristics in a given
organism and across species. An important subset of CAPs is the silent information regulator
(Sir) group. These proteins are involved in the silencing of genes in Saccharomyces cerevisiae
2
that is more commonly known as bakers or budding yeast. Saccharomyces cerevisiae, one of the
most widely used model systems, is an essential research tool for the understanding of both
genetic and epigenetic events (regulatory events important for regulation of gene expression that
do not involve changes in DNA sequences). One of the Sir family members, Sir3, is an
architectural CAP that was initially characterized as a histone-tail binding protein. However, it
was recently demonstrated to also have a DNA-binding ability (Georgel et al., 2001). Another
CAP family known to be involved in eukaryotic silencing is the methyl-DNA-binding protein
(MDB) group. Amongst this family of CAPs, the methyl-DNA binding protein 2 (MeCP2) is of
particular importance because of its links with the Rett syndrome (RTT) disorder (Amir et al.,
1999). RTT is a progressive childhood neurodevelopmental disorder that is one of the most
common causes of mental disabilities in white Caucasian female patients with an incidence rate
of 1 in 15,000 (Hagberg, 1985). Mutations in MeCP2 have been directly linked with the majority
of patients classified with RTT (Amir et al., 1999). This CAP was initially described as a methyl-
DNA binding global transcriptional repressor, but was recently demonstrated to also associate
with transcriptionally active promoters (Nan et al., 1997, Yasui et al. 2007; Chahrour et al.
2008). Mutations linked to RTT have also been found outside of the methyl-DNA binding
domain (MBD), indicating additional domains are necessary for proper MeCP2 developmental or
regulatory functioning (Amir et al., 1999). In addition to its methyl DNA binding role, MeCP2
has been described as a chromatin associated protein (Wade and Wolffe). A systematic in vitro
characterization of MeCP2’s role in chromatin dynamics described its ability to promote the
adoption of novel secondary structures suggesting that the function of MeCP2 in silencing gene
expression may be related to its ability to induce large-scale chromatin reorganization (Georgel
et al., 2003). These two CAPs appear to affect chromatin through multiple types of interactions,
3
and a more thorough investigation is required to more fully understand the mode of action of
these architectural CAPs.
The focus of this study was to characterize the chromatin binding determinants of the
CAPs Sir3 and MeCP2 and to clarify their role in chromatin architecture. Multiple studies have
characterized the CAP interaction with chromatin through a single determinant, however the
additional and related functions of these proteins have yet to be fully explained. Emerging
evidence seems to link multiple CAPs with more than one activity or target (Georgel et al. 2001;
Nielsen et al., 2001; Vakoc et al., 2005; Kanno et al., 1995; Jacobs et al., 1999; Molofsky et al.,
2003). The studies described herein are aimed at understanding the DNA-binding component of
Sir3, identifying and analyzing the preferences in binding to DNA features. These DNA
interactions will provide a better understanding of the complex role of Sir3 in silencing gene
expression. Similarly, we also report that MeCP2, in addition to recognizing methylated CpGs,
influences chromatin folding through histone N-Termini (or tails) interactions. Importantly,
MeCP2 mutants linked to the onset of RTT have the ability to differentially recognize and bind
specific histone tails. By further understanding the complex interplay between CAPs and
chromatin, and their impacts on gene expression, we may be able to manipulate these regulatory
events in order to treat genetic-linked diseases.
CHROMATIN COMPOSITION AND ORGANIZATION
In eukaryotic nuclei, DNA is packaged into a nucleo-protein complex referred to as
chromatin. This structure is highly complex and yet organized in a manner that still allows
sufficient DNA accessibility for such processes as transcription, replication, and DNA repair to
take place. The packaging of eukaryotic DNA requires increasingly complex levels of chromatin
condensation. The open structure, called euchromatin, is found in more highly transcribed
4
regions of the genome, while highly condensed regions, called heterochromatin, is found in
telomeres, centromeres, and non-coding sequences. Access to DNA regulatory elements is
achieved by the alteration of DNA-histone contacts that allows mobilization of nucleosomes
revealing the target sequences of DNA that are required to interact with transcription factors.
The structure of DNA was solved over 50 years ago, yet alterations in chromatin structure,
composition, and function directed by regulatory CAP- interactions remain to be fully elucidated.
The first level of DNA compaction results from its formation of a complex with an
octamer of DNA-binding proteins called core histones that form a nucleosome. The nucleosome
core particle (NCP) is the basic unit of chromatin and consists of 146 base pairs (bp) of DNA
wrapped in a left-handed manner around a histone core octamer comprised of two copies of each
core histone; H2A, H2B, H3, and H4 (Figure 1.1). Under physiological ion concentrations, DNA
is a helical structure that makes a complete turn every 3.4 nm (10.5 bp/turn). Nucleosomal DNA
is wrapped around the histone core octamer through 1.75 helical turns with a 10.2 bp/turn
average (Wolffe 1998). The curvature and shape of the DNA around the histone core octamer is
not uniform. Two regions of 10.0 bp/turn of DNA are known to flank the dyad axis of
nucleosomal DNA. The dyad axis region has a periodicity of 10.7 bp/turn. Severe DNA
distortion is found at the junction of these regions leading to a change in the average bp/turn
within nucleosomal DNA (Wolffe 1998). These distortions in the natural linear B-DNA type
structure have the potential to be recognized as a particular substrate or binding site, and
influence the recruitment of various DNA associated proteins and enzymes. For instance, the
preference of HIV integrase activity for nucleosomal DNA is due to the particular DNA structure
adopted while incorporated around core histones and is a useful example of how DNA structure
within chromatin can affect a biological process (Pruss et al., 1994). Nucleosome assembly has
5
been shown to be sequential, starting with the recruitment of a tetramer of histones H3 and H4,
followed by the addition of two H2A-H2B dimers. Each core histone has a structured domain
referred to as a histone fold that is comprised of three alpha helices (one long helix flanked by
two shorter helices) which allow histone-histone interactions (the alpha helices of each core
histone structure and interactions can be seen in the NCP crystal structure (Figure 1.1). The
unstructured N-terminal sections of core histones (N-terminal tail) are known to be structural
modulators of chromatin architecture, and serve as targets for PTMs. These histone N-terminal
tail PTMs can play a role in chromatin architectural changes, and influence CAP recruitment.
The DNA residing between two adjacent nucleosomes is referred to as linker DNA, and is a
short stretch of linear DNA that is minimally hindered sterically by the nucleosomal structure. A
collection of nucleosomes on a stretch of DNA is referred to as an array of nucleosomes, or
nucleosomal arrays (NA) (Wolffe 1998). An extended nucleosomal array has a diameter of 10-
nm, often referred to as beads-on a-string, and has recently been termed as the “primary”
structure of chromatin in a nomenclature analogous to that which describes levels of protein
folding (Woodcock and Dimitrov, 2001).
6
Figure 1.1: X-ray Crystal Structure of the Nucleosome Core Particle (NCP) *
*Modified from Luger et al, 1997
The secondary structure of chromatin results from condensation of the 10-nm primary
structure into a higher order 30-nm chromatin fiber. Despite extensive studies, there are still
large gaps in understanding this chromatin folding process, and the exact structure of the 30-nm
fiber is presently being debated. There are two proposed models of the 30-nm chromatin
structure; the solenoid chromatin model (Figure 1.2, A) (Finch and Klug, 1976; Thomas et al.,
1979; Felsenfeld & McGhee, 1986) in which a one-start helical stack of nucleosomes has bent
linker DNA between adjacent nucleosomes and the zigzag two-start helix model (Figure 1.2, B
and C) which proposes a relatively straight segment of linker DNA that connects two helical
stacks of nucleosomes (Figure 1.2; linker DNA highlighted in yellow to allow orientation within
the chromatin fiber) (reviewed in Wu et al., 2007). Within the zigzag model there have been two
possible structures proposed with different linker DNA conformations. The helical/twisted-
7
ribbon model (Figure 1.2, B) has the linker DNA orientated at angles ranging from 0° to 50°
(Worcel et al, 1981; Woodcock et al, 1984) and the crossed-linker model that has the linker DNA
perpendicular to the chromatin fiber axis (Williams et al, 1986; Smith et al, 1990). Recent
evidence seems to favor the zigzag two-start helix over the one-start solenoid model, but more
studies must be done to fully understand the condensation of nucleosomal arrays into the 30-nm
fiber (Dorigo et al, 2004; Kruithof et al, 2009). Studies involving in vitro reconstituted arrays
from purified components have demonstrated that the core histone octamer itself and the N-
terminal tails play a critical role in the folding of NA into the 30-nm structure (Fletcher and
Hansen 1995; Luger et al, 1997; Dorigo et al, 2004; Zheng et al., 2005; Kan et al., 2007; Kan et
al., 2009). Another set of DNA-binding proteins, linker histones H1/H5, are of particular interest
because they are found in nearly stochiometric abundance to nucleosomes and are known to
stabilize the intrinsic 30-nm secondary structure adopted by nucleosomal arrays under
physiological conditions (reviewed in Lugar and Hansen 2005). Another chromatin-associated
protein, heterochromatin protein 1 (HP1), was shown to associate and remodel chromatin
secondary structure through a series of in vitro biophysical studies and in vivo fluorescence
recovery after photobleaching (FRAP) analysis (Fan et al, 2004; Cheutin et al, 2003). After the
nucleosome, the 30-nm secondary chromatin structure has been the most intensively studied
structure in the field of chromatin biology, and yet the structural mechanics and interacting
architectural proteins are just now starting to come into focus.
8
Figure 1.2: Models of the 30-nm Chromatin Fiber * (A) Solenoid model. (B) Helical/Twisted-ribbon model. (C) Crossed-linker model.
* Modified from Wu et al. (2007)
There is currently very little known about the structure-function relationship of chromatin
compaction beyond the 30-nm diameter chromatin fiber. It is known that, due to the size of the
nucleus, nucleosomal arrays have to be folded into compact, higher order secondary and tertiary
structures. This intensive chromatin folding is achieved through the contribution of both histone
proteins and non-histone architectural proteins (Woodcock and Dimitrov, 2001; Luger and
Hansen, 2005). Supporting evidence comes from in vitro studies using reconstituted NA
showing that certain specific CAPs such as Myeloid and Erythroid Nuclear Termination stage-
specific protein (MENT), polycomb group protein complex (PcG), MeCP2, and Sir3 can
strongly influence chromatin compaction (Springhetti et al., 2003; Georgel et al., 2001; Georgel
et al., 2003 Francis et al., 2004). The exact structure and function of the higher order chromatin
architecture above the 30-nm fiber is currently mostly unknown, but this field of research is
critical considering that the level and orderly modulation of chromatin folding is intricately
linked to transcriptional control and developmental regulation (Wolffe, 1998).
9
EPIGENETIC MODIFICATIONS
The term epigenetics was first used in the late 1930s by Dr. Conrad Hal Waddington in
reference to what he described as the epigenetic landscape, a concept he proposed to explain how
the external manifestation of genetic activity influenced the cells fate (tissue types). Historically,
this arrived at a time when the individual components of the nuclei were still thought to be
unimportant, and only after the discovery of DNA as heritable information was his research was
revisited and seriously considered. In the 1970s, studies began to demonstrate that not only was
the sequence of DNA influencing phenotypes, but another layer of control was dictating genetic
events (Holliday and Pugh, 1975; Weintraub and Groudine, 1976; Gottesfeld and Butler, 1977;
Lohr et al., 1977). The regulation of gene expression through steric hindrance of transcription
factor recruitment on promoter regions by both DNA methylation and nucleosome location
ushered in an age of what is now known as epigenetic research. Subsequent studies have
demonstrated additional epigenetic effects associated with certain RNA molecules (non-coding
RNA), histone PTMs, and CAPs (Figure 1.3).
10
Figure 1.3: Different Mechanisms of Epigenetic Regulation *
* Modified from Probst et al. (2009)
DNA-METHYLATION
Methylation of DNA was first suggested as an important epigenetic mechanism that
could control gene activity in higher organisms by Drs. Holliday and Pugh in 1975 (Holliday and
Pugh, 1975). Today, DNA methylation is known to be present in a wide range of organisms, both
prokaryotic and eukaryotic, and operates by two vastly different mechanisms. In prokaryotes,
DNA methylation occurs on both adenine and cytosine bases and this modification is used as a
mechanism of defense in which the host restriction system can differentiate itself from foreign
11
sources of DNA. This ensures that foreign DNA (un-methylated) can be degraded without any
deleterious effects to the bacterial DNA. In mammals, DNA methylation has only been reported
on cytosine bases, mainly in the context of CG dinucleotides or CNG trinucleotides. The
methylation of the eukaryotic genome has been linked to gene repression and found to co-
localize with condensed chromatin structures (Bird and Wolffe, 1999). Cytosine methylation
occurs on the carbon-5 position of a cytosine base through the transfer of a methyl group from S-
adenosylmethionine (SAM) by DNA methyltransferases (DNMT) (Figure 1.4). This epigenetic
modification is vital for proper eukaryotic development as evidenced by the lethality of the
DNMT-null mutation in mice (Li et al.., 1992; Okano et al., 1999). Recently, the nucleotide 5-
hydroxymethylcytosine has been discovered and described to be located in relatively high
abundance in the brain (Kriaucionis and Heintz 2009). This altered form of DNA methylation
has been proposed to participate in the regulation of neuronal genes, but the epigenetic
mechanism of this DNA modification is currently unknown.
Figure 1.4: Mechanism of DNA Methylation *
SAM= S-Adenosyl Methionine; DNMT= DNA methyltransferase
* Modified from Expert Reviews in Molecular Medicine © Cambridge University Press (2002)
It is estimated that 80% of CpGs are methylated in the DNA of somatic cells, and the
majority of unmethylated CpGs reside in specific areas called CpG islands. This pattern of
12
genomic DNA methylation and subsequent gene inhibition is inherited by replicating daughter
cells, allowing tissue specific gene methylation and repression. In contrast housekeeping genes
with CpG islands located in their promoter regions are expressed ubiquitously across cell types
(Figure 1.5, A). The density and location of methyl-CpGs in the promoter of genes are known to
be critical for gene expression levels (Boyes and Bird 1992; Hsieh 1994).
Figure 1.5: DNA Methylation * HDAC= histone deacetylase; MBD= methyl-DNA binding domain
* Modified from Bergman and Cedar (2004)
DNA-methylation inhibits gene expression through one of three mechanisms in
eukaryotes: 1) direct inhibition of transcription factors/DNA interactions modulated by
methylation of specific regions of promoters (Bell and Felsenfeld 2000; Hark et al, 2000; Szabo
et al, 2000; Holmgren et al, 2001), 2) recognition of methylated DNA sequences by
transcriptional repressors associated with the recruitment of co-repressors (Jones et al, 1998; Nan
et al, 1998; Zhang et al, 1999), or 3) association of methylated-DNA binding proteins with
nucleosomes leading to chromatin compaction (Georgel et al, 2003). The first mechanism of
DNA-methylation inhibition physically blocks transcription factors’ recognition of their cognate
DNA sequences, whereas the second and third mechanisms rely on recognition of the
methylated-DNA or specific chromatin feature(s) as binding sites for methylated DNA binding
13
proteins (MBDs). Gene repression from the second mechanism operates through MBPs coverage
of methylated loci hindering transcription factors access to DNA sequences or elements. These
MBPs can also recruit additional co-repressors to areas of DNA methylation (Figure 1.5, B), and
can further the repression of genes through enzymatically modifying local chromatin (Nan et al.,
1998). Through the third mechanism, other methylated-DNA binding CAPs can directly
condense nucleosomes into a higher order chromatin structure known to be repressive to
transcriptional machinery. These three silencing mechanisms may be further modulated, through
a currently undetermined mechanism, by the newly discovered 5-hydroxymethylcytosine
discussed previously (Kriaucionis and Heintz 2009). This DNA modification’s function in gene
regulation is currently unknown, but it is likely that the change in structure of the methyl group
by the addition of a hydroxyl group could affect all three of these methylation-binding linked
mechanisms.
HISTONE POST-TRANSLATIONAL MODIFICATIONS
The unstructured N-terminal regions of core histones are known to be subjected to
various PTMs that can affect chromatin structure and function. These chromatin changes can
occur through mechanisms that alter nucleosome stability and subsequent access to nucleosomal
DNA or by signaling for CAPs recruitment. The CAPs themselves can affect chromatin stability
and/or structure (Luger and Hansen, 2005). Of these histone PTMs, acetylation and methylation
are the most thoroughly studied and are found to be intricately involved in the regulation of gene
activity (Table 1.1). In addition to these two PTMs, histone phosphorylation, ubiquitination,
ADP ribosylation, deimination, isomerization, and SUMOylation have been described. These N-
terminal tail PTMs can serve as recognition and binding sites for a variety of CAPs mediated by
14
interactions with specific domains. The discovery of these signals has lead researchers to propose
a “histone code” or “chromatin language” that pertains to regulation of transcription associated
with specific PTMs (Allis, 2001; Berger, 2007). In addition to signaling, the PTMs of histone
tails and core regions can interfere with DNA-histone contacts by altering residue charges,
exposing sequences or elements of DNA that would otherwise be sterically blocked.
Table 1.1: Post-Translational Modifications Effects on Transcription *
* Modified from Berger (2007)
Acetylation and methylation of lysines are the most studied and best characterized of all
the nucleosomal PTMs. Acetylation of core histones is mainly associated with regions of actively
transcribed genes and a more open chromatin structure. Lysine acetylation is carried out by the
enzyme family of histone acetyltransferases (HATs) and is directed at the N-terminal tails of all
four core histones and on a few sites within the core globular domain. In contrast, histone
deacetylases (HDACs) are enzymes that remove this PTM, and are recruited to silent or
repressed loci. Acetylation of histones is known to influence chromatin through two separate
means: recruitment of CAPs by specialized domains that recognize acetylated lysines, called
15
bromodomains, and physical disruption of the histone-DNA electrostatic interaction by the
reduction of positively charged residues that interact with the negatively charged phosphate
backbone of the DNA. Whereas lysine acetylation is linked to active transcription, lysine
methylation can be involved in either activation or repression of genes depending upon the
residues modified (Table 1.1). The arginine and lysine residues of histones can exist in one of
three states of methylation: monomethylation, dimethylation, or trimethyalation. These three
forms of histone methylation can be selectively recognized by CAPs and therefore be involved in
differential signaling. Histone methylation was once believed to be a nonreversible reaction but
recent studies have identified the first histone demethylase enzyme, LSD1 (Shi et al., 2004).
Since then at least ten additional enzymes have subsequently been characterized as histone
demethylases (reviewed in Anand and Marmorstein, 2007). The effect of histone methylation on
chromatin structure is “enforced” through the recruitment of chromodomain-containing CAPs.
Chromodomains are motifs that can specifically recognize methylated lysines within histone
tails. These chromodomains are found in proteins that are known to either directly remodel
chromatin, or be incorporated into macromolecular chromatin structures. For example, di- or tri-
methylation of histone H3 lysine 9 (H3K9me) is a well-characterized repression signal that is
frequently referred to as a hallmark of heterochromatin. The role of H3K9me in repression was
confirmed by immunofluorescence localization with high enrichment found in heterochromatin
regions (Rice et al., 2003). The protein HP1, a CAP mentioned previously, is a known
chromodomain protein that specifically recognizes H3K9 di-methylation and is found to co-
localize with this PTM (Lachner et al., 2001; James et al., 1989; Nakayama et al., 2001). HP1
was initially characterized as a heterochromatin associated transcriptional repressor (reviewed in
Kwon and Workman 2008), but has since been observed in euchromatic areas of the genome
16
where it can complex with either silenced genes (Nielsen et al., 2001) or actively transcribed loci
(Vakoc et al., 2005). This association with actively transcribed genes is evidence for either a
separate role for HP1 in gene regulation or an example of chromatin higher order structure
playing a role in gene transcription. There is currently little information on the recognition of
arginine methylation by CAPs. Arginine methylation has only been linked to transcriptional
activation because of evidence that the modification localizes on nucleosomes in active genes
(Boisvert et al., 2005). One recent study has shown the methylation of arginine 2 on histone 3
(H3R2me) hinders CAP binding to H3 tails (Iberg et al., 2008). This study combined with the
lack of identified proteins with methylated arginine specific domains indicates a possible role for
this PTM in signaling through CAP binding impediment. Histone methylation is increasingly
found to be one of the more complex PTMs whose exact role in epigenetic regulation is still
under investigation.
Histone phosphorylation (Table 1.1) is a histone PTM that has been linked to a wide
range of nuclear functions. This PTM is known to be involved in the DNA repair pathway,
transcriptional regulation, and the mitotic chromatin condensation process. The mechanism of
phosphorylation in gene activation is not fully understood. It is believed that this PTM hinders
histone tail-DNA contact by the conjugation of a phosphate group which adds negative charges
to serine or theronine residues. This reduction in the basic charge of histone tails is believed to
reduce its affinity to the negatively charged phosphate backbone of DNA. In addition to altering
histone-DNA contacts, the presence of phosphorylation on histone H3 serine 10 (H3S10)
stimulates HAT activity on the histone tail and is believed to increase transcription through this
histone-modification crosstalk (Cheung et al., 2000; Lo et al., 2000; Clayton et al., 2000).
Phosphorylation of histones is also found on nucleosomes at sites of double-stranded DNA break
17
occurrences, and is believed to be a key signal in the DNA repair pathways (Madigan et al.,
2002; Nakamura et al., 2004; Fernandez-Capetillo et al., 2004). Furthermore, histone H3
phosphorylation is known to play a role in heterochromatin formation during the chromatin
condensation that occurs in mitotic division (Guo et al., 1995; Ajiro et al., 1996). In addition to
core histone modification, the phosphorylation of linker histones is a key regulatory event of
gene expression. This PTM stimulates the release of the linker histone by altering electrostatic
interactions between residues and DNA. The release of linker histones destabilizes the higher
order structure of condensed chromatin (Kaplan et al., 1984; Lin et al., 1991). Protein
phosphorylation has long been known to be a regulatory PTM signal and now is known to play
multiple roles in nuclear processes through nucleosome alteration and CAP signaling.
The other core histones PTMs are not as well characterized. The role of lysine
ubiquitination of histones H2A and H2B remains ill-defined, but may be involved in
transcription initiation, silencing, and DNA repair (Wang et al., 2004; Ikura et al., 2007; Zhou et
al., 2008; Nakagawa et al., 2008; Zhao et al., 2008). Ubiquitination is similar to phosphorylation
in its proposed involvement in gene activation and DNA repair. These two PTM also have a
large number of both transferases (>500 protein kinases and >600 ubiquitin ligases) that can
catalyze these modifications, and hydrolases (>150 protein phosphatases and >100
deubiquitinating enzymes) that reverse them (Hunter, 2007). ADP ribosylation has been
identified as a modification to each core histone, but its function/s is currently unclear (Burzio et
al., 1979). A few studies have shown that this PTM is involved in the DNA repair process and
cell proliferation (Kreimeyer et al., 1984; Boulikas, 1989; Boulikas, 1990). Early architectural
studies using ADP-ribosylated chromatin as a substrate demonstrated a possible role in
chromatin folding. ADP-ribosylated templates display a fairly open configuration, with the
18
removal of ADP-ribose resulting in a more condensed state (Frechette et al., 1985; de Murcia et
al., 1986). Some researchers have even hypothesized that the ADP ribosylation of histones could
be involved in crosstalk with NAD(+)-dependent pathways and might be a direct link between
the regulation of bioenergetics and gene transcription (reviewed in Hassa et al., 2006). Histone
SUMOylation occurs through the conjugation of SUMO, a small ubiquitin-like protein, to lysine
residues in all four core histones (Nathan et al., 2006). SUMOylation of histones has been
demonstrated to antagonize histone acetylation linking this PTM with gene repression (Nathan et
al., 2006). The PTM isomerization refers to the two conformational states peptidyl proline can
adopt and is a non-covalent histone PTM. The difference between these two conformational
states is the dihedral angle of the peptide bond between the proline and the preceding residue that
differs by 180o, dramatically altering the polypeptides secondary structure. This PTM is involved
in both repression and activation of transcription through the mechanism of these two
conformational states acting as a differential signal for additional histone modifying enzymes
(Nelson et al., 2006). Deimination as a PTM is the process of converting an arginine into a
citrulline residue, an amino acid that is not directly coded for by DNA. This histone deimination
is known to antagonize arginine methylation based transcriptional induction (Cuthbert et al.,
2004).
HISTONE VARIANTS
In addition to the canonical core histones, variant histones have been discovered which
have specialized functions relating to DNA repair, replication, and transcription (Table 1.2).
These variants can be incorporated into the histone core octamer, replacing one of the four core
histones. This change in NCP composition introduces an additional level of complexity when
examining the relationship of chromatin with gene regulation and nuclear events. Histone
19
variants alter chromatin dynamics by having distinct N-terminal tails comprised of an altered
amino acid sequence that allow them to obtain cell regulatory PTMs which are not compatible
with the four common core histones (reviewed in Bernstein and Hake, 2006). In addition to these
altered N-terminal signaling capabilities, some of these histone variants display variations in
their globular domain that can cause structural and functional alterations in nucleosome and
chromatin structure. These structural changes can be recognized by specific CAPs. One recent
study showed that the CAP, HP1, was more effective at modulating chromatin if the histone
variant H2A.Z was incorporated into an array of nucleosomes (Fan et al, 2004). This example of
a histone variant regulating CAP interactions was the first evidence linking a histone variant
incorporation to a chromatin architectural protein’s function (reviewed in Luger and Hansen,
2005). H2A.Z, often associated with boundary regions of heterochromatin in Saccharomyces
cerevisiae, is believed to block the spread of heterochromatic domains, but the exact mode of
action is still under investigation (Meneghini et al., 2003; Xu et al., 2005). The PTMs of histone
variants can also be critical in the regulation of nuclear events in a cell. Phosphorylation of the
histone H2A variant, H2A.X, is one of the most well characterized modifications of histones and
is a key signal in the repair pathway of DNA double-stranded breaks (Rogakou et al., 1998;
Arkady et al., 2003). There are numerous histone variants with a wide range of biological
functions (Table 1.1) that have been identified across species, but are beyond the scope of this
thesis. The interactions between histones, histone variants, and CAPs are now known to be a
critical component in the epigenetic regulation of cellular activity.
20
Table 1.2: Function and Location of Variant Histones * TS= Tissue Specific; RD= Replication Dependent; RI= Replication Independent; ND= Not Determined; TG= Throughout Genome; a= DNA Damaged Induced; b= Found at Borders of Heterochromatin/Euchromatin in Saccharomyces cerevisiae,
excluded from mammalian Xi; Xi= Inactive X Chromosome
* Modified from Bernstein and Hake (2006)
CHROMATIN-ASSOCIATED PROTEINS
The packaging and maintenance of DNA within the chromatin environment is performed
by a broad range of proteins termed chromatin-associated proteins (CAPs). One of the least
understood aspects of cellular functioning is how DNA is packaged into the tight confines of the
nucleus. This compaction is highly dynamic as certain genomic loci remain accessible and
functionally active, while other areas are repressively compacted. These CAPs are typically
divided into functional or structural groups based on enzymatic activity, histone-like
structure/properties, architectural contribution, or transcriptional regulatory activity. Amongst
21
the most studied and best characterized groups of CAPs are those who display an enzymatic
activity. These enzymes are generally involved in modifying both histone and non-histone
components of chromatin. Histone modifying enzymes, such as histone acetyltransferases,
methyltransferases, and phosphokinases have been previously defined in this dissertation (see
page 13). In addition to these types of histone modifying enzymes another important group of
enzymatic CAPs are key components of the ATP-dependent chromatin remodeling factors.
These ATP-based complexes have also been shown to modulate nucleosome positioning and/or
composition directly affecting chromatin structure. This ATPase-based chromatin remodeling
facilitates rapid chromatin rearrangement inside the nucleus that occurs in response to outside
stimuli or developmental signals. Additional CAPs that do not contain an enzymatic function but
play a role in chromatin architecture have been identified. These architectural CAPs are involved
in the condensation of DNA and nucleosomal arrays into higher order chromatin structures. The
involvement of binding sites, enzymatic activities, and chromatin modulation capabilities of
CAPs has been shown to be critical components in gene expression regulation and genome
maintenance. Disruptions of these CAP’s functions have been linked to a wide range of diseases
including cancer and neurodevelopmental disorders. The identification of these associations has
led to an increased effort in the research community to characterize the regulation, activity, and
interacting partners of CAPs.
The two best characterized ATP-dependent chromatin remodeling families are the
SWI/SNF and ISWI-containing complexes (reviewed in Racki and Narlikar, 2008). These two
families are classified based on the structure of their catalytic subunit. Both of these ATPase
families use the energy of ATP hydrolysis to translocate DNA from and around histone
octamers, but they are known to operate by different mechanisms. The SWI/SNF complexes
22
hydrolyze ATP to remove or slide nucleosomes, alter histone octamer composition by exchange
of dimers, or create nucleosomes with DNA loops (Phelan et al., 2000; Yang et al., 2007; Bazett-
Jones et al., 1999). In contrast, the ISWI complex has only been shown to slide nucleosomes,
altering their positioning but not removing or exchanging core histones (Corona et al., 1999).
Initially the SWI/SNF complex was labeled a transcriptional activator due its enzymatic activity
leading to an increase in DNA accessibility (Hirschhorn et al., 1992; Peterson and Herskowitz
1992). The resulting “free” DNA was believed to become a better target for transcription factor
recruitment to DNA regulatory elements, promoting transcription (Imbalzano et al., 1994; Kwon
et al., 1994). These findings led to the subsequent classification of chromatin remodelers as
transcriptional activators. It has since been demonstrated that ATP-dependent chromatin
remodelers can affect both activation and repression of gene transcription through alterations of
chromatin structure (Trouche et al., 1997; Zhang et al., 1998; Murphy et al., 1999). The functions
and regulation of ATP-dependent chromatin remodelers is the current focus of ongoing research
at numerous laboratories.
The architectural function of CAPs is mostly characterized at the histone-DNA level of
interaction. This primary level of chromatin compaction has been studied extensively over the
past 30 to 40 years and was addressed in an earlier section of this dissertation (see page 4), but
far less is known about the higher-order arrangement of chromatin. Recent biochemical and
biophysical studies with non-histone CAPs have demonstrated mechanisms leading to re-
arrangements of NA secondary and tertiary structures (reviewed in Lugar and Hansen, 2005).
Non-histone architectural CAPs such as MENT, PcG proteins, HP1α, and MeCP2 have been the
focus of some recent in vitro chromatin folding studies. These studies have provided a few
insights into higher order structure (Springhetti et al., 2003; Francis et al., 2004; Nielson et al.,
23
2001; Georgel et al., 2003). Out of these architectural CAPs, MeCP2’s effect on chromatin
condensation has been characterized at the highest structural resolution (Georgel et al., 2003).
MeCP2 structure and function will be more fully described in a later section of this dissertation.
The MENT protein was initially identified as a major component of chicken granulocyte
heterochromatin (Grigoryev and Woodcock 1998). It has since been further characterized for its
role in condensing chromatin into unique secondary and tertiary chromatin structures
(Springhetti et al., 2003). One of the regions identified as responsible for this chromatin
condensation is MENTs M-loop domain. This region is known to contain a nuclear localization
signal, an AT-hook motif, and a DNA-binding domain first identified in the high mobility group
(HMG) proteins. This in vitro study also established the serpin (serine protease inhibitor) domain
or reactive center loop (RCL) to promote MENTs self-oligomerization. This RCL region had
previously been shown in vivo to be involved in MENT’s chromatin interactions (Irving et al.,
2002). Through RCL deletion mutants, this oligomerization ability of MENT was demonstrated
to be responsible for the majority its intra-chromatin fiber interaction activity (Springhetti et al.,
2003). This finding led researchers to propose that MENT condenses chromatin through two
separate mechanisms. The first is the condensation of nucleosomes within a chromatin fiber
through its M-loop domain, and the second operating through the formation of protein “bridges”
between chromatin fibers.
The PcG proteins were first identified in Drosophila melanogaster and were believed to
repress gene transcription by creating condensed heterochromatic structures (Denell 1973;
Messmer et al., 1992). In Drosophila, PcG proteins are known to repress the HOX genes whose
silencing is necessary to preserve the body patterning of developing flies (Lewis, 1978; Struhl,
1981; Simon et al., 1992). In mammals, PcG proteins have been characterized to be involved in
24
cell cycle regulation, cancer genesis, and stem cell self-renewal (Kanno et al., 1995; Jacobs et al.,
1999; Molofsky et al., 2003). This CAP has recently been shown by electron microscopy (EM)
to condense model chromatin fibers into compact secondary chromatin structures (Francis et al.,
2004). This compaction into higher order chromatin structures is theorized to be part of the
mechanism of gene silencing that occurs in vivo. This chromatin compaction induced by PcG
proteins was demonstrated to operate through interactions with nucleosomes. Tailless arrays
were also analyzed and the N-termini tails of histones were not found to be involved in PcG
mediated chromatin condensation (Francis et al., 2004).
The heterochromatin protein, HP1, was also first identified in Drosophila melanogaster,
and was found to be a dose-dependent transcriptional repressor associated with the position
effect variegation (PEV) silencing effect of heterochromatic regions (James and Elgin, 1986;
Eissenberg et al., 1990). It is probably the best characterized of all the known non-histone CAPs.
HP1 has long been described as one of the critical CAPs responsible for the creation and
maintenance of heterochromatic regions in a wide range of eukaryotic organisms. The one
notable exception is Saccharomyces cerevisiae, or budding yeast, which uses the Sir proteins for
PEV silencing. The Sir proteins will be discussed in more detail in a following section. HP1 has
been found to specifically recognize, bind, and co-localize with the methylated histone H3K9
and H3K27 heterochromatic marks (Bannister et al., 2001; Lachner et al., 2001). HP1 transcripts
are differentially spliced and exist as three different isoforms, HP1α, HP1β, and HP1γ, in mice,
humans, and Drosophila. These isoforms have a similar amino acid sequence and domain
organization, but differences reside in their cellular localization and interacting partners (Minc et
al., 1999). HP1 contains two characterized domains: a chromodomain (CHD) at the N-terminus
that is responsible for its heterochromatin binding, and a chromoshadow domain (CSD) at the C-
25
terminus that is involved in HP1 dimerization and in other protein interactions. These two
domains are separated by a flexible hinge region that is highly variable in amino acid sequence
between isoforms and organisms and contains multiple sites for PTMs (Badugu et al., 2005;
Koike et al., 2000). The majority of HP1 structural data has been obtained through biophysical
examination of condensed model chromatin fibers (Fan et al., 2004). This HP1α in vitro
structural data (analytical ultracentrifugation sedimentation velocity) demonstrated a requirement
for the histone H4 N-termini to allow compaction of chromatin fibers.
Another well-characterized family of CAPs is the HMG proteins. This protein super-
family is comprised of 3 HMG sub-groups: HMGA, HMGB, and HMGN. Each of the HMG
family members contains different chromatin binding motifs, but are all known to modulate
chromatin fiber properties (reviewed in Gerlitz et al., 2009). HMGA proteins, formerly known as
HMG-I/Y/C, are ubiquitously found in mammalian nuclei, and their function has been linked
with gene regulation, cell differentiation, and carcinogenesis (Landsman and Bustin, 1991;
Berlingieri et al., 1995; Chiappetta et al., 1995; Liberati et al., 1998). The characteristic feature
of these CAPs is an AT-hook domain involved in chromatin interactions (Reeves and Nissen,
1990). HMGA is known to be a highly disordered protein with little to no secondary structure,
but it can undergo disorder-to-order structural changes upon binding to DNA and/or other
proteins (Reeves and Wolffe, 1996; Huth et al., 1997). The HMGB protein family, formerly
called HMG-1/2, is characterized by its HMG-Box domains. HMGB contains two HMG-box
domains, HMG-box A and HMG-box B, which are this CAP family’s major DNA/chromatin
interaction sites. In addition to these two domains, HMGB’s C-terminal tail is acidic and
mediates additional DNA interactions (Lee and Thomas, 2000; Watson et al., 2007). The HMGB
protein family has been found to play a role in both transcription and replication, but can also act
26
as an extracellular signaling molecule (Landsman and Bustin, 1991; Pil and Lippard, 1992;
Agresti et al., 2005; Topalova et al., 2008). The last family, HMGN, previously referred to as
HMG-14/-17, is the only HMG family known to directly bind to nucleosomes, and the
nucleosomal binding domain (NBD) of HMBN is its characteristic feature (Crippa et al., 1992).
In addition to the NBD domain, the C-terminal region of HMGN contributes to its chromatin
binding, as truncated mutants display a significant decrease in binding affinity (Ueda et al.,
2004). HMGN proteins have been shown to affect chromatin higher order structure and
influence the expression of a subset of genes (Lim et al., 2005; Zhu and Hansen, 2007; Belova et
al., 2008).
EUKARYOTIC GENE SILENCING
Silencing in eukaryotic organisms is generally associated with transcriptionally
repressive, compact heterochromatic regions of the genome. In the budding yeast,
Saccharomyces cerevisiae, silenced regions are found in three genomic areas; the silent mating
type loci, the repetitive ribosomal DNA (rDNA) genes, and the telomeres. While the components
and composition of heterochromatin in Saccharomyces cerevisiae varies from other eukaryotes,
many of the molecular mechanisms of heterochromatin silencing are conserved. In higher
eukaryotes, heterochromatin is found at telomeres and centromeres, and is known to be involved
in the proper segregation of chromosomes during replication. Similar to yeast, the increased level
of chromatin compaction found in the heterochromatin of higher eukaryotes is refractory to
transcription and replication. In Drosophila melanogaster, heterochromatin is known to
contribute to the position effect variegation that influences the transcription of genes depending
on their location and proximity. In mammals, heterochromatin is additionally associated with the
X-chromosome inactivation found in females. One significant difference amongst organism’s
27
genomes is the level of DNA methylation present. DNA methylation of promoters in higher
eukaryotes plays a significant role in regulating gene transcription. Abnormal gene silencing can
lead to a host of complications for an organism, with cancer in humans being a key example.
Understanding the components and molecular mechanism of eukaryotic gene silencing is
therefore essential to developing treatments and cures for a wide range of diseases.
SIR FAMILY OF PROTEINS
Silencing in Saccharomyces cerevisiae is heavily dependent on the Sir family of proteins.
The functions and structures of the Sir proteins vary widely. However, they all contribute to the
establishment and/or maintenance of silent chromatin in yeast. The Sir1 protein is not considered
critical for silencing, but it does aid in the assembly of other Sir proteins, which enhances the
establishment of heterochromatin (Pillus and Rine 2006). The Sir2 protein belongs to another
larger family of proteins that have been found to be NAD+-dependent deacetylases (Smith et al.,
2000, Imai et al., 2000). This histone deacetylase function of Sir2 is not required for the
initiation of silencing, but it is important to note that hypoacetylated histones were described to
preferentially recruit Sir3, leading to an increase in silencing (Suka et al., 2001; Carmen et al.,
2002). Recent evidence also indicates a structural role for the product of Sir2’s enzymatic
reaction, O-ADP-ribose, in Sir-mediated heterochromatin formation (Martino et al., 2009). The
family member Sir4 has no known enzymatic activity, but is believed to be responsible for
protein-protein interactions that are involved in the recruitment and association of Sir2, and to a
lesser extent Sir3, to silenced areas (reviewed in Rusche et al., 2003).
28
SILENCING INFORMATION REGULATOR 3 (SIR3)
The silencing protein Sir3 is known to be a critical architectural protein required for the
spreading and maintenance of silencing in Saccharomyces cerevisiae (Renauld et al., 1993,
Carmen et al., 2002). Silencing of genes due to the genomic proximity to telomeric and silent
mating type loci is well established in current literature (Laurenson and Rine 1992; Sandell and
Zakian 1992; Renauld et al., 1993; Barton and Kaback 2006). The increase in distance of a gene
from a telomere also leads to a direct decrease in the frequency of silencing of a telomere-
adjacent gene (Renauld et al., 1993). Sir3 has been characterized as the only limiting protein for
the propagation of telomeric silencing and this spread of heterochromatic silencing has been
linked to Sir3 dosage indicating the importance of this silencing protein in transcriptional control
(Renauld et al., 1993). Additional studies using sir 3 null mutant yeast have confirmed Sir3 as
critical in the position-effect of silent loci (Carmen et al., 2002). Genes that exhibit telomere
position effect have increased transcription levels with removal of the adjacent telomere, and
similar results were found in sir3-null mutants, further galvanizing Sir3’s role in the spread of
telomeric silencing (Barton and Kaback, 2006). The previously proposed model of spreading of
telomeric silencing has been described to be initiated with the DNA-sequence specific binding of
repressor activator protein 1 (Rap1), causing recruitment of the Sir3 and Sir4 proteins to the
telomeric regions (reviewed in Rusche et al., 2003). This heterochromatin complex formation
was further described as the result of multiple weak interactions that take place between the Sir
proteins and the N-termini-tails of histone H3 and H4 (Hecht et al., 1995). Though Sir3 has long
been described as interacting with/and silencing chromatin solely through these histone tail-
interactions, the DNA binding ability of Sir3 has been demonstrated in vitro (Georgel et al.,
2001). This newly discovered Sir3 property has been proposed to be important for the spreading
29
of silencing at specific loci, leading to at least a partial re-evaluation of the actual role of Sir3
and possibly Sir4 in the establishment and maintenance of silent chromatin domains.
Figure 1.6: Sir3 Domain and Partner Interaction Location *
* Modified from Stone and Pillus (1998)
The identification of Sir3’s functional domains has increased the understanding of its role
in regulating nuclear events. Sir3 is known to interact with an array of nuclear proteins that are
involved in different functions ranging from DNA repair (Rad7) to transcriptional silencing
(Rap1, Histones H3 and H4) (Shore et al., 1984; Palladino et al., 1984; Moretti et al., 1994;
Hecht et al., 1995). Sir3’s C-terminal region has been the focus of numerous studies related to its
role in heterochromatin formation. The interaction of Sir3 with the histone tails of H3 and H4
was linked to residues residing at Sir3’s C-terminus end (Bell et al., 1995). These interactions
have classically been described to be the primary mechanism for Sir3-recruitment and
heterochromatin formation.
30
The ability of Sir3 to oligomerize has also been linked to the C-terminal region and
recently two separate oligomerization domains within this area have been identified (Liaw et al.,
2006; King et al., 2006). This ability of Sir3 has now been further characterized through
analytical ultracentrifugation studies (McBryant et al., 2006). The ability to self-associate has
been proposed to affect chromatin through a nucleosome “bridging” mechanism that promotes
interactions with both nearby nucleosomes and more widely spaced nucleosomes (reviewed in
McBryant et al., 2006). The C-terminal region of Sir3 has also been linked to DNA repair
through its interactions with Rad7 (Paetkau et al., 1994) and Rad52 (Park et al., 1999). The C-
terminal structured area consists of the binding sites for Rap1, Rad7, and the N-termini of
histones H3 and H4. The Sir3 protein also contains an AAA domain that in other members of
the AAA domain family couples the hydrolysis of ATP to conformational changes that aid in
protein complex assembly/disassembly (recently reviewed in Erzberger and Berger 2006). In
Sir3, the AAA domain lacks the critical residues that are known to be responsible for ATP
binding. It had been theorized that the Sir3 AAA domain could bind to 2’-O-acetyl-ADP ribose
which is a product of the Sir2 deacetylation reaction. Recent in vitro evidence supports this
theory by demonstrating O-acetyl-ADP ribose promotes the formation of Sir3 mediated
heterochromatic complexes (Martino et al., 2009).
The N-terminal region of Sir3 has been the focus of many studies due to this region
seeming to have multiple functions. The amino acids 1-214 on the N-terminal end of Sir3 show
50% identity to the largest subunit of the origin of recognition complex, Orc1, in a domain
suspected to be responsible for its transcriptional silencing function (Bell et al., 1995). The
bromo-adjacent homology (BAH) domain of Sir3 is located in the first 214 residues, and is
sufficient to silence in the absence of a full-length protein but only with over-expression of Sir1
31
(Connelly et al., 2006). The function of the BAH domain has been predominantly characterized
by its role in Orc1 mediated transcriptional silencing (Bell et al., 1995), and has been implicated
in the protein-protein interactions of Orc1 and Sir1 (Hou et al., 2005). It is believed that Sir3’s
BAH domain may interact with Sir1 and this protein-protein contact may be involved in Sir3’s
recruitment. The BAH domain of Sir3 has also recently been described as a nucleosome/histone
tail binding domain that interacts with the H4 N-terminal tail and H3 globular domain (Onishi et
al., 2007). Deletion of the first 235 AA of N-terminal region of Sir3 abated its ability to function
in mating-type repression in vivo, demonstrating the importance of this domain in Sir3
localization (Bell et al., 1995). The N-terminal region of Sir3 (1-503 AA), alone, when over-
expressed was sufficient to increase the frequency and area of telomere-proximal silencing, a
phenotype similar to that obtained with full length Sir3 over-expression (Gotta et al., 1998). It is
not known if this phenotype results from the increased interaction with components of the
silencing machinery, or if this truncated mutant is releasing full-length Sir3 from the complexes
increasing the pool of available fully functioning protein. The Sir3 BAH domain structure has
recently been determined to a 1.9 A resolution (Connelly et al., 2006). The crystallized Sir3
BAH domain formed mostly a β-sheet with a helical H domain. The H domain of Orc1 is known
to interact with the Sir1 C-terminal domain. Through circular dichroism (CD), one third of Sir3
has been shown to be disordered with the majority residing in a 300-amino acid stretch between
the structured N and C terminals (McBryant et al., 2006). The structured N-terminal region of
Sir3 corresponds with the BAH domain.
While Sir3 does not possess a classic DNA-binding motif, it does contain patches of
positively charged residues that may be responsible for Sir3’s interaction with the DNA through
the negatively charged phosphate backbone. While regions of Sir3 have been studied and shown
32
to be important for silencing, the individual contributions of the different functions of Sir3 in
particular the role of Sir3’s DNA binding interaction on heterochromatin formation are not yet
fully understood.
POST-TRANSLATIONAL MODIFICATIONS OF SIR3
There are a number of studies indicating the phosphorylation of Sir3 can influence its
silencing function. Initial experiments demonstrated Sir3 became hyperphosphorylated under
heat shock and starvation, and in response to mating pheromone (Stone and Pillus 1996). This
Sir3 hyperphosphorylation was linked with an increase in transcriptional silencing at telomeric
regions. Another study demonstrated through inhibiting Sir3 phosphorylation, that the PTM
status of Sir3 affected its subtelomeric silencing More recent studies have identified the
phosphorylation of the serine residue 275 (S275) of Sir3 by the Slt2p MAP kinase pathway (Ray
et al., 2003). This S275 phosphorylation was also shown to redistribute Sir3 silencing, and
decrease the yeast’s lifespan upon commitment to cell growth. The phosphorylation of Sir3 is
believed to be a possible post-translational event that could serve as a trigger mechanism for
changes in Sir3-mediated silencing.
METHYLATED-DNA BINDING PROTEIN FAMILY
Because links between gene silencing and DNA methylation have been demonstrated, the
cellular mechanism of methylated DNA recognition by methyl-DNA binding domain proteins
(MBD) has emerged as an important research focus in gene regulation. The MeCP2 protein was
the first MBD protein discovered to selectively recognize and bind methylated DNA sequences
(Lewis et al., 1992; Meehan et al., 1992). Since the discovery of MeCP2, four additional
members of the MBD family (MBD1, MBD2, MBD3, and MBD4) have been identified through
33
polypeptide sequence bioinformatic analysis of the methylated-DNA binding domain (MBD)
shared by these proteins (Figure 1.7) (Hendrich and Bird 1998). In addition to the common MBD
motif, MBD1, MBD2, and MeCP2 contain a similar C-terminal transcription repression domain
(TRD). This TRD has been identified as the region involved in the gene repression activity
associated with these CAPs (Nan et al., 1997). The most recently identified methyl-DNA binding
protein (MBP), Kaiso, lacks the characteristic MBD found in other MBPs (Table 1-N)
(Prokhortchouk et al., 2001). This MBP is atypical in its method of methylated-DNA binding by
utilizing a zinc finger (ZF) domain for DNA methylation recognition. Instead of a TRD, Kaiso
has a poxvirus and zinc finger/Bric-a-brac, Tramtrack, Broad-complex (POZ/BTB) domain that
is responsible for its transcriptional repression function.
Figure 1.7: The Known Members of the Mammalian methyl-CpG-binding Protein Family * POZ= poxvirus and zinc finger domain; BTB= Bric-a-brac, Tramtrack, Broad-complex domain; MBD= methyl-binding domain;
ZF= zinc-finger domain; CxxC= zinc-binding domains; TRD= transcriptional repression domain; GR= 11 repeat stretch of glycine/arginine residues
* Modified from Klose and Bird (2006)
The mechanism of MBD silencing was previously discussed in the DNA methylation
section of this dissertation. Briefly, the MBD family of proteins is believed to recognize and
selectively bind to methylated regions of the genome. Through this selective binding, these
34
MBPs have classically been described as repressors of transcription through steric masking of
DNA elements and/or recruitment of chromatin remodeling co-repressor complexes to DNA
methylated loci. Out of all the MBD proteins, MeCP2 has been the most thoroughly studied due
to its link with the neurodevelopmental disorder RTT, which will be discussed more thoroughly
in a later section.
Additional functions of the MBD family outside of MeCP2 have recently been
elucidated. For instance, the MBD1 protein has been found to form a stable complex with a H3
lysine 9 (H3K9) methyltransferase (SETB1) (Sarraf et al., 2004). This complex interacts with the
replication machinery during the S phase of the cell cycle and was observed to facilitate H3K9
methylation during replication-dependent chromatin assembly. Also, the MBD2 protein has
been linked to genetic silencing through methylation specific binding over gene promoters
(Hendrich et al., 2003). MBD2 is the methyl-DNA-binding component of the methyl-CpG-
binding protein 1 (MeCP1) transcriptional repressor complex (Ng et al., 1999). Prior to MeCP2’s
discovery, the MeCP1 complex was characterized to bind to methylated DNA regions, but its
individual peptide composition and their contribution to the complexes function was unclear at
the time (Meehan et al., 1989). The MeCP1 complex has since been characterized in vitro to
preferentially bind methylated nucleosomes, deacetylate histone N-termini, and remodel
nucleosome positioning (Feng and Zhang, 2001). The MBD2 protein has also been shown to be
critical for tumorigenesis in mouse intestinal cancer, indicating a possible role for this MBD in
the regulation of cell proliferation (Sansom et al., 2003).
The MBD3 protein is associated with the MBP family based on the MBD sequence
homology, but it lacks key residues in the MBD motif responsible for methylated DNA
interaction (Hendrich and Bird 1998). Recently, MBD3 has been identified as a component of
35
the nucleosome remodeling and deacetylation (NuRD) co-repressor complex, a well
characterized transcriptional repressor involved with genetic silencing in a range of organisms
(Wade et al., 1999; Denslow and Wade 2007). The MBD3 protein is not capable of selectively
recognizing DNA methylation, but is required for proper NuRD formation/stabilization (Kaji et
al., 2006). MBD3 is also important for development as an Mbd3 null mutation was found to be
embryonically lethal in mice (Hendrich et al., 2001). The MBD4 protein has been mostly
characterized for its role in DNA repair (Hendrich et al., 1999; Millar et al., 2002; Wong et al.,
2002), but has also recently been implicated in transcriptional silencing (Kondo et al., 2005). The
C-terminal glycosylase domain of MBD4 is not found in any other MBD proteins, and is
essential for its DNA repair function (Hendrich et al., 1999).
The Kaiso protein represses transcription of methylated genes in Xenopus laevis through
its association with the nuclear receptor co-repressor (N-CoR) complex which possesses a
histone deacetylase activity (Yoon et al., 2003). Kaiso null-mutations in Xenopus laevis clearly
show up-regulation of genes with methylated promoters (Ruzov et al., 2004), but this phenotype
was not apparent in similar mouse studies (Prokhortchouk et al., 2006). The depletion of this
MBP in mice resulted in a resistance to intestinal tumorigenesis similar to that described in
MBD2-null experiments (Prokhortchouk et al., 2006), but additional studies are needed to further
characterize Kaiso’s role in mammals.
METHYL-CpG-BINDING PROTEIN 2 (MeCP2)
As previously stated, MeCP2 was the first of the MBD family of proteins to be identified
(Lewis et al., 1992; Meehan et al., 1992). The MBD motif characterized in the MeCP2 protein
was used to ascertain the other MBD family members (including the MBD2 protein, a sub-unit
36
of the MeCP1 complex); consequently MeCP2 has been termed the “founding member” of the
MBD (or MBP) protein family. MeCP2 was also the first MBP found to interact with HDAC-
containing complexes, linking two epigenetic repression mechanisms: DNA methylation and
histone deacetylation (Nan et al., 1998). Biological interest in MeCP2 rose exponentially with
the discovery of a genetic link between MeCP2 and the neurodevelopmental disorder RTT (Amir
et al., 1999). The vast majority of patients with RTT have been found to harbor alterations in
either MeCP2 or protein sequence, leading researchers to propose this protein’s function to be
critical for normal neurodevelopment. In support of this theory, additional mutations of the
mecp2 gene have recently been linked with other neurodevelopmental disorders such as autism
and Angelman-like syndrome (Watson et al., 2001; Hammer et al., 2002; Shibayama et al., 2003;
Carney et al., 2003).
MeCP2 PATHOLOGY (RTT AND CANCER)
The Rett Syndrome (RTT) was originally described by Andreas Rett in a German article
in 1966, but this work was not extensively distributed across the world’s medical community.
The RTT disorder was more widely known after an English publication by Dr. Hagberg in 1983
(Hagberg et al., 1983). This childhood neurodevelopmental disease’s pathology is characterized
by normal development for up to 6-18 months followed by a period of slowed development
occurring towards the end of this period. This period of slowed progress in development is
followed by regression that affects mostly the patient’s speech and hand coordination. Later,
RTT characteristics manifest in more advanced, stereotypical hand gestures, such as hand
wringing. In addition, patients may develop an abnormal gait, autistic behavior, and a reduction
in growth. The severity of RTT disability in patients can vary widely, with the type of Mecp2
mutation somewhat influencing the disease’s course and development. RTT was initially
37
described to affect mostly Caucasian females with an incidence of 1 in 15,000 births (Hagberg
1985). The understanding of its molecular mechanism was advanced enormously when the RTT-
associated locus was mapped to the region of the X-chromosome that contained the Mecp2 gene
(Amir et al., 1999). The majority of RTT patients are now known to have mutations in the Mecp2
gene, however patient diagnosis is currently based on clinical observations and not by genetic
testing. Clinical studies have now demonstrated that this syndrome can also affect male patients,
but at a much lower rate due to the X chromosome link (Meloni et al., 2000; Orrico et al., 2000;
Van Esch et al., 2005). Although when RTT is found in male patients, its severity is much
greater, since all of the male’s cells have only one copy of the gene which would be the mutant
form (Meloni et al., 2000; Couvert et al., 2001). The molecular mechanism of RTT is only
partially understood, and there is currently no known cure or treatment.
An additional role for MeCP2 during cancer genesis has recently been described. The
hypermethylation of tumor suppressor gene promoters is a well characterized event in cancer
genesis (Merlo et al., 1995; Gonzalez-Zulueta et al., 1995). The initial evidence of MeCP2
involvement with cancer arose when it was found that the methylation of the breast cancer 1
gene (BRCA1) promoter in the presence of MeCP2 resulted in repression. Hypermethylation of
additional tumor-suppressor genes in cancer have also been shown to be associated with MeCP2
and other MBDs (Wischnewski et al., 2007; Ballestar et al. 2003; Bakker et al. 2002; Magdinier
et al., 2001). Studies involving prostate cells demonstrated MeCP2 to be critical in both normal
and cancerous cell growth (Bernard et al., 2006). This study, through the ectopic expression of
MeCP2, showed that in androgen-dependent prostate cancer cells, MeCP2 promotes growth
without androgen stimulation. These prostate cancer cells were also able to sustain tumorigenic
properties during androgen depletion indicating a role for MeCP2 in repressing tumor
38
suppressing genes. Interestingly, individual mouse knock outs for MBD1, MBD2, or MeCP2 did
not affect tumor formation, suggesting an overlapping function for these MBDs in tumor
suppression (Hendrich et al., 2001; Zhao et al., 2003). Further characterization of MeCP2s role
in cancer genesis is needed to elucidate its individual function in tumor suppression.
ISOFORMS OF MeCP2
There are two alternatively spliced isoforms of Mecp2 transcripts (Kriaucioniz and Bird,
2004; Mnatzakanian et al., 2004). These isoforms, MeCP2e1 and MeCP2e2, differ only in their
N-terminal regions (Figure 1.6). The spliced variant MeCP2e1 has a 21 residue section with an
acidic pI of 4.25, whereas the e2 isoform of MeCP2 has only 9 residues over the same section
that possesses a basic pI of 9.5 (Mnatzakanian et al., 2004). It is interesting to note that the
MeCP2e2 isoform was the first identified variant of MeCP2 and therefore most characterized,
but the MeCP2e1 isoform is more abundant in the brains of both mice and humans
(Mnatzakanian et al., 2004). These two isoforms have further been characterized to have
differential distribution in developing postnatal mouse brains between the dorsal thalamus and
hypothalamus (Dragich et al., 2007). However, recent studies have shown both of these isoforms
to co-localize to heterochromatic regions in murine fibroblast cells (Kumar et al., 2009). Though
differences in structure and distribution have been demonstrated, there are currently no known
functional differences described between these two MeCP2 isoforms.
FUNCTION AND STRUCTURE OF MeCP2
MeCP2 has classically been described and mostly characterized as a global
transcriptional repressor (Nan et al., 1997). After the identification of the methyl-DNA binding
MeCP1 complex from crude nuclear extract, MeCP2 was characterized as the second
39
nucleoprotein that could bind methylated-CpGs (Lewis et al., 1992). This MeCP2 protein, unlike
the MeCP1 complex, could bind to DNA that contained only a single methylated CpG (Lewis et
al., 1992). The ability of MeCP2 to bind to methylated DNA combined with its localization to
murine pericentromeric heterochromatin initially led researchers to propose this protein to be a
methyl-DNA binding transcriptional repressor (Lewis et al., 1992). Since this discovery, the
majority of studies on MeCP2’s function have attempted to elucidate its role in this repression
mechanism. Another early study on MeCP2 revealed that it was over 100 times more abundant
than MeCP1 in the nucleus, and it could bind to either methylated or unmethylated DNA to
repress transcription (Meehan et al., 1992). In contrast to earlier experiments, additional in vitro
gene expression studies indicated that MeCP2 only repressed promoters that contained CG
methylation, and slightly enhanced transcription levels were reported when MeCP2 was bound to
unmethylated promoters (Nan et al., 1997). This study also described MeCP2 to contain not only
a MBD, but also a TRD that alone can repress the transcription of genes (Nan et al., 1997).
Regions responsible for non-specific DNA interaction within MeCP2 have been identified in
both the MBD and TRD (McBryant et al., 2007), although full-length MeCP2 has been found in
vitro to preferentially bind methylated over unmethylated DNA templates (Ballestar et al., 2000).
In addition to these two repression-linked domains, MeCP2 is known to further inhibit gene
activity through interactions with histone modifying enzymes. MeCP2 was characterized early
on to form a complex with Sin3 and HDAC (Nan et al., 1998). This association is theorized to
further repress transcription at the MeCP2 recognized methylated loci through the HDAC histone
deacetylase activity. MeCP2 has also been shown to interact with an unidentified histone
methyltransferase through immuno-precipitation experiments (Fuks et al., 2003). This study
showed an increased localization of MeCP2 correlated to an increase in the amount of H3K9
40
methylation at the observed loci. In vitro chromatin binding studies using chromatin as template
and purified human MeCP2 protein demonstrated this CAP to possess, in addition to its other
functions, a chromatin condensation ability (Georgel et al., 2003). This chromatin condensation
was accompanied by the formation of higher order oligomeric structures at increased MeCP2
molar ratios. The MeCP2-mediated higher order structures were observed to contain intra-
chromatin fiber interactions in addition to inter-chromatin condensation. This study also
demonstrated, by using unmethylated chromatin, that this MeCP2-mediated chromatin
compaction function was separate from its methyl-DNA binding preference. In mice, MeCP2
null-mutations displayed a loss of long-range chromatin interactions, providing further
indications that MeCP2 regulates chromatin structure in vivo (Horike et al., 2005). The recent
discovery of an architectural role for MeCP2 in higher order chromatin structures is not fully
characterized, and is a major focus in this dissertation.
Recently, researchers have unexpectedly discovered MeCP2 to have additional functions
related to RNA processing (Young et al., 2005) and active transcription (Yasui et al., 2007;
Chahrour et al., 2008). The association with RNA splicing was indicated initially through a Co-
IP mass spectrophotometric identification of an interaction between MeCP2 and an RNA binding
protein: the Y-box binding protein 1(YB-1) (Young et al., 2005). The same study demonstrated
that MeCP2 itself can bind to RNA and regulate splicing in vivo. In addition to its RNA splicing
activity, MeCP2 has also been found to associate with actively transcribed genes (Yasui et al.,
2007; Chahrour et al., 2008). Through the use of chromatin immunoprecipitation-microarray
chip technology (ChIP-chip assays), MeCP2-binding sites on 26.3 megabases of imprinted and
non-imprinted neuronal cells were analyzed (Yasui et al., 2007). This study found 59% of
MeCP2 binding sites resided outside of known genes, and unexpectedly, 63% of the MeCP2-
41
bound promoters were transcriptionally active. Additional mouse studies using hypothalamus
that either lacked or overexpressed MeCP2 demonstrated that ~85% of affected genes were
activated when higher MeCP2 levels were detected (Chahrour et al., 2008). This MeCP2-induced
activation finding was confirmed by the identification of MeCP2 on the promoters of six selected
activated genes. In addition, MeCP2 was also found to associate with the cAMP responsible
element binding protein 1 (CREB1) transcriptional activator on an active, but not a repressed
form of the same gene. These studies demonstrate that the cellular role of MeCP2 is more
complex than simply repressing transcription through its methyl-DNA binding ability and
association with co-repressors.
Structural studies of MeCP2 have yielded few insights into its cellular functions.
Classically, only two domains have been associated with MeCP2 function. The one structural
domain that has been solved is the MBD (Figure 1.8) (Wakefield et al., 1999), which is the same
domain that was used to define this family of CAPs (Hendrich and Bird 1998). This MBD of
MeCP2 is 63 residues in length and consists of four anti-parallel beta sheets that generate a
wedge shaped structure. On one side of the beta sheet wedge, the two longer beta sheets are
believed to interact with the major groove of the DNA where the methylation specificity would
take place. The TRD has been characterized through truncated MeCP2/Gal4 DNA binding
domain fusion products (Nan et al., 1997). The area between amino acid 205 through 310 was
found to be sufficient to repress transcription of target genes. The mechanism of TRD repression
is currently unknown. The C-terminal end of MeCP2, beyond the TRD, has also been implicated
as an important contributor in RTT as indicated by the number of identified truncated mutants in
diagnosed patients (Amir et al., 1999). This C-terminal region has recently been described as
containing a WW binding domain important for MeCP2s interactions with splicing factors
42
(Buschdorf and Strätling 2004). WW binding domains are known to recognize proline residues
of interacting ligands. WW domains are characterized by the signature presence of two
tryptophan residues (W) that are separated by 20-22 amino acids. In addition to these three
characterized regions, biophysical and protease digestion experiments have identified 6 distinct
domains (Adams et al., 2007). Two domains identified corresponded with the previously
described locations of the MBD and TRD domains. The HMGD1 and HMGD2 domains of
MeCP2 share sequence similarities with the previously mentioned high mobility group AT-hook
2 (HMGA2) protein. Furthermore, the trypsin digestion analysis of MeCP2 sub-divided the CTD
into a CTD α and a CTD β. The CTD β domain contains the WW binding domain which is
involved in splicing factor interactions and also contains a stretch of 7 consecutive histidines
between residues 366 and 372. There is currently no function attributed to the CTD α domain.
Figure 1.8: The Described Domains of the Methyl-CpG-binding Protein 2 (MeCP2) * HMGD1= HMG (High Mobility Group protein)-like domain 1; MBD= methyl-binding domain; HMGD2= HMG (High Mobility Group protein)-like domain 2; TRD= transcriptional repression domain; CTDα= C-terminal Domain alpha; CTDβ= C-terminal
Domain beta
* Modified from Georgel et al. (2003) and Hite et al. (2009)
The tertiary structure of MeCP2 has been characterized to be atypical for nucleoproteins,
as it contains a large amount of intrinsic disorder. The calculated mass of a MeCP2 monomer
43
based on amino acid composition is 53 kDa, but gel filtration experiments yield a molecular
mass of 500 kDa, and SDS-PAGE migration yields an apparent size of 75-80 kDa (Klose and
Bird 2004). To date, there is no NMR or X-ray crystal structure of the full length MeCP2 protein.
Biophysical and biochemical methods have been used to study its unusual tertiary structure. CD
analysis of full-length MeCP2 demonstrated almost 60% of the protein to be unstructured
(Adams et al., 2007). Of the structured portions, ~35% were β-strand/turns and ~5% were α-
helices. Through analytical ultracentrifugation, MeCP2 has been shown to exist as a monomer
across a wide range of ionic conditions and molar concentrations (Adams et al., 2007). MeCP2
also displayed an unusually low sedimentation coefficient of 2.2 S and a correspondingly high
frictional coefficient ratio (f/fo = 2.4). The results of the CD and analytical ultracentrifuge
experiments suggest MeCP2 has a coil-like tertiary structure that is similar to a partially
denatured protein. Though some progress has been made, full elucidation of the influence of
MeCP2 structure on function is critical for understanding the molecular link between mutations
and the development of RTT.
POST-TRANSLATIONAL MODIFICATION OF MeCP2
The post-translational modification status of MeCP2 has recently been found to modulate
its gene regulatory activities (Zhou et al., 2006; Tao et al., 2009). Phosphorylation of the serine
421 (S421) residue was the first PTM of MeCP2 to be described, and was found to be induced
during increased neuronal activity associated with influx of calcium ions. This phosphorylation
event was therefore proposed to be triggered by a CaMKII (Ca2+/calmodulin-dependent protein
kinase II)-reliant mechanism. Out of twelve tissue types tested, this PTM was found only on the
MeCP2 protein from brain tissue, suggesting phosphorylation of S421 to be solely a neuronal
event (Zhou et al., 2006). The phosphorylation of S421 was further characterized in these studies
44
to be involved in alleviating the MeCP2-regulated transcriptional repression of the Bdnf (brain-
derived neurotrophic factor) gene. These findings were the first to demonstrate a direct
correlation between a PTM of MeCP2 and its gene regulatory activities. Further studies have
identified additional MeCP2 phosphorylation sites from the brain samples of multiple species
(Tao et al., 2009). The MeCP2 serine 80 (S80) was one of these residues characterized as
phosphorylated through immunoblotting experiments. MeCP2 S80 phosphorylation was found to
be dependent on the neuronal calcium influx, similar to what was demonstrated for S421.
Although in contrast to the previous S421 characterization, the S80 phosphorylation was
negatively regulated by neuronal activity, suggesting alternative signaling between these residues
in resting or depolarized neurons. The phosphorylation of both S421 and S80 was found to
decrease the association of MeCP2 at specific loci. Interestingly, MeCP2 binding to gene
promoters was not always indicative of gene repression, supporting the alternative role for
MeCP2 as a co-activator. An additional four MeCP2 phosphorylation sites have been identified
(Tao et al., 2009), but the occurrence and function of these PTMs and any cumulative effect
remains to be characterized. The potential effect of phosphorylation on MeCP2s secondary or
tertiary structures has not yet been investigated.
MeCP2 RTT MUTATIONS
Genetic analysis has identified 218 mutations linked to patients diagnosed with RTT
(Miltenberger-Miltenyi and Lacone, 2003). Mutations have been found in all six MeCP2-
characterized domains, indicating that each individual domain contributes to proper function.
These mutations are usually single point mutations in the coding sequences that result in a
missense or nonsense mutant, altering a single amino acid or causing a truncated form of the
protein. Almost all the cases reported to date have been sporadic mutations to the Mecp2 gene
45
with limited familial exceptions. Eight of these Mecp2 mutations have been most commonly
found in RTT patients and have been termed mutational “hot spots” (Figure 1.9) (Kriaucionis
and Bird, 2003). Interestingly, the majority of these mutations occur at arginine residues. The
MeCP2 mutations linked with RTT have been the focus of numerous studies aiming at
discovering the molecular mechanism of this neurodevelopmental disease. While MeCP2s
function has mainly been attributed to its methylation specificity, two of the more common RTT
mutations the T158M and the R168X mutants do not affect the MBD function. The T158M point
mutant resides within the MBD, but has been demonstrated to have only a modest twofold
decrease in affinity for methylated DNA as opposed to the more drastic reduction in other RTT
MBD mutants (Ballestar et al., 2000). In addition the R168X truncated mutant has an intact, fully
functional MBD. These two common mutations, as well as the additional truncated and non-
MBD point mutants identified, indicate MeCP2 functions other than methyl-CpG DNA binding
is critical for normal neurodevelopment. After the chromatin condensation ability of MeCP2 was
demonstrated, the effect of the RTT R133C and R168X mutations on chromatin interactions was
examined (Georgel et al., 2003). The MeCP2 R133C mutant had previously been shown to lose
its DNA methylation binding specificity (Ballestar et al., 2000), and was further characterized in
an additional study to have no significant difference in its ability to interact with chromatin
(Georgel et al., 2003). However, the R168X mutant lost the ability to form higher order
chromatin oligomeric structures. Further characterization of the functional differences between
wild type MeCP2 and RTT mutants is needed and is a focus of this dissertation. Identifying the
structural role of MeCP2 in the dynamic chromatin folding process will contribute to the
46
elucidation of the molecular mechanism of RTT syndrome.
Figure 1.9: Location and Frequency of MeCP2 “Hotspot” Mutations in RTT Patients * R106W= Arginine to Tryptophan Point Mutation at Residue 106; R133C = Arginine to Cysteine Point Mutation at Residue 133;
T158M= Threonine to Methionine Point mutation at Residue 158; R168X= Arginine to Stop Codon at Residue 168; R255X= Arginine to Stop Codon at Residue 255; R294X= Arginine to Stop Codon at Residue 294; R306C= Arginine to Cysteine Point
Mutation at Residue 306
* Modified from Zlatanova (2005) and Hite et al. (2009)
47
CHAPTER 2: ROLE OF NUCLEIC ACID BINDING IN SIR3p-DEPENDENT INTERACTIONS WITH CHROMATIN FIBERS
Nicholas L. Adkins1, Steve J. McBryant,2, Cotteka N. Johnson,1, Jennifer M. Leidy, 1,3,
Christopher L. Woodcock,4, Charles H. Robert,5, Jeffrey C. Hansen,2, and Philippe T. Georgel1.
AUTHOR ADDRESS: Department of Biological Sciences and Cell Differentiation and
Development Center, Marshall University1, West Virginia 25701, Department of Biochemistry
and Molecular Biology, Colorado State University2, Fort Collins, CO 80523-1870, School of
Medicine, West Virginia University3, Morgantown, WV 26506-9100, CNRS - Institut de
Department of Biology, University of Massachusetts, Amherst4, Amherst, MA 01003
Biochimie et Biophysique Moléculaire et Cellulaire5, Paris Sud, Orsay 91405 France,
AUTHOR EMAIL ADDRESS: [email protected]
TITLE RUNNING HEAD: Nucleic acid binding of Sir3.
CORRESPONDING AUTHOR FOOTNOTE: Address correspondence to: Philippe T. Georgel,
PhD, Marshall University, Department of Biological Sciences, 1 John Marshall Drive,
Huntington WV 25755. Fax; 304-696-7136.
Published by Biochemistry 2009 Jan 20; 48(2):276-88.
48
ABSTRACT
Recent studies of the mechanisms involved in the regulation of gene expression in
eukaryotic organisms depict a highly complex process requiring a coordinated rearrangement of
numerous molecules to mediate DNA accessibility. Silencing in Saccharomyces cerevisiae
involves the Sir family of proteins. Sir3p, originally described as repressing key areas of the
yeast genome through interactions with the tails of histones H3 and H4, appears to have
additional roles in that process, including involvement with a DNA binding component. Our in
vitro studies focused on the characterization of Sir3p-nucleic acid interactions and their
biological functions in Sir3p-mediated silencing using binding assays, EM imaging, and
theoretical modeling. Our results suggest that the initial Sir3p recruitment is partially DNA-
driven, highly cooperative, and dependent on nucleosomal features other than histone tails. The
initial step appears to be rapidly followed by the spreading of silencing using linker DNA as a
track.
49
INTRODUCTION
The regulation of gene expression in eukaryotes is dependent on several sets of proteins
that regulate DNA accessibility. In this context, the level of chromatin compaction plays a
central role in the transition from a transcriptionally active to a repressed state. In yeast, gene
silencing requires the establishment of a specific chromatin configuration that displays a high
level of similarity to the heterochromatin conformation observed in higher eukaryotes.
Saccharomyces cerevisiae possesses a set of specialized proteins called Silent Information
Repressors (Sir 1-4) (Rine and Herskowitz, 1987) that, in combination with ORC1 and RAP1,
promote nucleation and spreading of silencing at specific loci (Loo and Rine, 1994). Sir 2 is a
NAD-dependent histone deacetylase (Rine and Herskowitz, 1987), and Sir3p and Sir4p are
structural chromatin-associated proteins initially described as N-terminus binding partners for
histones H3 and H4 (Loo and Rine, 1994; Hecht et al., 1995; Hecht et al., 1996; Moazed et al.,
1997; Moretii and Shore, 2001). The mechanism of Sir-mediated silencing is intimately linked
with structural modifications of chromatin (Hecht et al., 1995; Hecht et al., 1996; Georgel et al.,
2001). The recruitment of Sir3 and Sir4 has initially been described as a process mediated by
their binding to the histone tails of H3 and H4 (Hecht et al., 1995; Hecht et al., 1996), although
biochemical data indicate that the tails are not needed for binding to chromatin in vitro (Luger et
al., 1997; McBryant et al., 2008). Recent publications have also indicated a potential role for
double-stranded DNA (dsDNA) in the establishment of silencing theorizing that the DNA itself
may provide a means for establishment and/or spreading of silencing (Georgel et al., 2001; Liaw
and Lustig, 2006). Our initial experiments suggested that Sir3p would bind more readily to the
linker DNA in chromatin than to DNA wrapped around a histone octamer, but in a different
manner than that of linker histone (Georgel et al., 2001). In addition, the Sir3p concentration was
50
shown to affect chromatin fiber-fiber interactions, resulting in potential large-scale chromatin re-
organization (Georgel et al., 2001).
The goal of this study is to provide a better delineation of the role of nucleic acids in the
Sir3p-mediated recruitment and formation of supra-molecular structures in vitro. To this end, we
have carried out experiments using normal and tailless nucleosome arrays as controls, as well as
nucleic acid fragments of various compositions, lengths, and conformations. The study also
includes analysis of Sir3p binding as a function of Sir3p and DNA concentrations, to mimic
possible changes in local concentrations within the nucleus.
To determine the effect of DNA conformation on Sir3p recruitment, we used several
short linear DNA fragments with various intrinsic bendabilities. Our results indicate that Sir3p
does not significantly differentiate between linear, curved, or bent DNA. However, a synthetic
four-way-junction (FWJ) designed to mimic the nucleosomal entry-exit DNA region (Georgel et
al., 1997) is a poorer substrate for DNA binding. Using a binding assay employing DNA
fragments of decreasing sizes, we determined that Sir3p binding occurs over templates as small
as 12 bp. Finally, we investigate the influence of Sir3p molarity on inter-molecular DNA
bridging using either DNA, nucleosomal arrays (NA), or trypsinized or tailless NAs as binding
templates, and find that an increased local concentration promotes DNA bundling. Electron
microscopy images are consistent with our electrophoretic mobility shift assay (EMSA) results
and demonstrate the presence of Sir3p-induced layered DNA fragments. In combination with our
observations that Sir3p can efficiently form complexes with tailless NA (Georgel et al., 2001;
McBryant et al., 2008), we conclude that the mechanism by which Sir3p induces gene silencing
is not only associated with histone tails, but is in part DNA-driven.
51
EXPERIMENTAL PROCEDURES
Sir3p purification- The protein was purified from SF9 cells infected with a recombinant
bacculovirus containing a recombinant Sir3p sequence fused to six C-terminal histidines
(Georgel et al., 2001, Bell et al., 1995). After 40 hours, nuclear extracts prepared from infected
cells were purified through a Ni affinity chromatography column followed by Q-Sepharose (Bell
et al., 1995). After purification, Sir3p was dialyzed against H5EGN (5 mM HEPES, 0.1 mM
EDTA, 5% glycerol, 10 mM KCL, 0.1% NP-40). The concentration was determined by
spectroscopy based on the extinction coefficient, and confirmed by comparison to BSA standards
after SDS-PAGE as described in Georgel et al. (Mumberg et al., 1995). Alternatively, Sir3p was
over-expressed and purified from a bacterial system as follows. Sir3p was expressed from pJC52
(Georgel et al., 2001) in BL21 cells (Stratagene). Expression was induced by addition of IPTG
to a final concentration of 1 mM for 5 hours in the presence of 3% ethanol. Sir3p was purified
using TALON Metal Affinity Resin (Clontech) according to the manufacturer’s specifications.
The protein was subsequently dialyzed against H5EGN, and the concentration was calculated by
comparing Coomassie blue staining against that of purified BSA, and confirmed by spectroscopy
based on the extinction coefficient. Note that both forms of Sir3p displayed indistinguishable
nucleic acid and chromatin binding properties when used for titration experiments.
DNA and nucleosomal arrays- The 208-7 and 208-12 DNA templates containing repeats
of the 5S rDNA from Lytechinus variegatus were purified from plasmids pPol I 208-7 and pPol I
208-12 (Georgel et al., 1993), after digestion with Hha1, followed by gel-filtration purification,
as described in Hansen et al. (Hansen et al., 1989). Histone octamers were purified from chicken
erythrocytes (Georgel et al., 2001), and tailless histones were prepared by limited trypsin
digestion as described by Fletcher and Hansen (Fletcher and Hansen, 1995). Nucleosome arrays
52
(normal or tailless) were reconstituted by salt dialysis at a ratio of 1.1 mole octamer: 1 mole 208-
bp of DNA (Hansen and Lohr, 1993). The efficiency of reconstitution was analyzed by
ultracentrifugation and Quantitative Agarose Gel Electrophoresis (QAGE), as described by
Fletcher et al. (Fletcher et al., 1994).
PCR fragments and oligomers used for gel electrophoresis analysis- The linear DNA
fragment referred to as “Straight” was prepared by PCR amplification using the plasmid pdHSP
XA 0.5 (Becker and Wu, 1992) as the template, and primers p1 and p2 encompassing positions -
185 to -165 (p1: 5’TCG AGA AAT TTC TCT GGC CG3’) and +18 to +36 (p2: 5’TTC GCG
ATG TGT TCA CCT3’) from the Drosophila hsp70 gene. The 199 bp “Bent” fragment
encompassing the Acanthamoeba histolitica RNA polymerase 1 binding site was generated from
the plasmid pPol I 208-12 (Georgel et al., 1993) using primers p3 (5’CGC TCG TTT TAC AAC
GTC3’) and p4 (5’CCG CAC AGA TGC GTA AGG3’).
The 208-1 5S rDNA was used as a source of material for the “Curved” fragment. It was
prepared by Ava1 digestion of the 208-12 DNA fragment (10 units/µg of DNA for 60 minutes at
37oC) and gel-purified after electrophoresis in a 1% agarose gel. The 100 bp fragment was
prepared by PCR amplification (primers). The 50 bp fragment was also PCR-amplified using
primers p5 (5’TCG ACG AAG CGC CTC T3’) and p6 (5’AGG CGC GCT CTC TCT C3’) and
the plasmid pdHSPXA 0.5 as template (Becker and Wu, 1992).
The 32 bp fragment was generated by annealing primers p7 (5’TTC AGG CGC GCG
CTA GCG AAG CAA CAG AG3’) and p8 (5’CTC TGT TGC TTC GCT AGC GCC CGC CTG
AA3’). The 12 bp fragment was generated similarly by annealing the primers p9 (5’TCA CTT
ATT TGT3’) and p10 (5’ACA AAT AAG TGA3’).
53
Evaluation of DNA curvature- The “Straight”, “Bent”, and “Curved” DNA conformations
were assessed by QAGE, the DNA fragments were electrophoresed in multigel as described in
Fletcher et al. (Hansen and Lohr, 1993). The effective radii (Re) were determined using the
following equation:
µ/µ'o= (1 - Re/Pe) 2
where µ represents the mobility of the fragments, µ 'o is the surface-charge density, Re is the
effective radius, and Pe is the pore size of the agarose gels [500 ng of DNA plus 500 ng of T3
phage (used as an internal marker)]. After 4 hours of electrophoresis at 1.33 V/cm in TAE buffer,
the multigels were stained with SYBR green for 30 minutes. The distances of migration of all
samples were measured and Pe was calculated from the information obtained from the
concomitant electrophoresis of the T3 phage internal marker, whose known size allows
calibration (Fletcher et al., 1994). The DNA conformations and overall curvatures were then
estimated based on their respective sizes and Re values. The theoretical calculations to determine
the conformation of the DNA fragments were performed using the Trifonov algorithm, as
described in Georgel and Robert (Georgel and Robert, 2002).
Reconstitution of a four-way-junction- The four-way-junction template was assembled
from four separate oligonucleotides by sequentially annealing equimolar amounts of two sub-sets
of single-stranded oligos at 65oC
(FWJ1: GATCCTAGGCCTCACGTATTATATCGATGCATGCG,
FWJ2: AATTCGACGATCGAAGCTGAATACGTGAGGCCTAGG,
FWJ3: ACCATGCTCGAGATTACGAGCAGCTTCGATCGTCG, and
FWJ4: TTCGCATGCATGCATCGATATCTCGTAATCTCGAGCATGG).
54
First, oligonucleotides FWJ2 and FWJ3 were incubated in standard TE buffer (10 mM Tris HCl
and 1 mM EDTA) at 65oC for 10 minutes, then slow-cooled (~ 1-2 degrees per minute) to 10oC
to allow the formation of DNA duplexes. Oligonucleotides FWJ1 and FWJ4 were treated
similarly. The two intermediate DNA products (duplexes) were mixed together, re-heated to
65oC and again slowly brought down to 10oC. The formation of the four-way junction structure
was monitored and confirmed by electrophoresis in a 10% non-denaturing acrylamide gel by
comparing the respective electrophoretic mobility of the single-stranded oligonucleotides, the
intermediate duplexes (FWJ2-FWJ3 and FWJ1-FWJ4), and the completed four-way junction, as
described by Panyutin et al. (Panyutin and Hsieh, 1993). Based on the measured intensity of
SYBR green staining, we estimated that 90% of the single-stranded oligonucleotides were
incorporated into four-way junctions (data not shown).
Sir3p-nucleic acid Electrophoretic Mobility Shift Assays (EMSA)- Sir3p titrations of
DNA templates were performed using Electrophoretic Mobility Shift Assays. A given quantity
of Sir3 protein was mixed with the DNA template and equilibrated at 25oC for 30 minutes. The
EMSA were performed using DNA at either 5 ng/µl or 10 ng/µl. The experiments were
compared using the same values of Sir3p molar ratios (rSir3, corresponding to moles of Sir3 to
moles of 208 bp of DNA, as defined in Georgel et al. (Georgel et al., 2001)). Aliquots (10 µl for
normal EMSA or 20 µl in competition assays) of the equilibrated solutions containing different
amounts of Sir3p were loaded onto 1% agarose gels (longer fragments: ~200 bp and above) or
6% acrylamide (shorter DNA fragments: 12-100 bp and single-stranded DNA) for quantification
of the free-DNA template band by integration of the Southern blot signal intensities (using NIH
ImageJ). In each quantified gel, control lanes containing DNA with no added Sir3 were loaded
55
and co-electrophoresed in order to provide internal calibration standards, which also provided an
estimate of the standard deviation of measured free-DNA values.
Electron microscopy- Samples were prepared for transmission electron microscopy
essentially as described (Nikitina et al., 2007). Briefly, DNA and Sir3p were mixed at the
desired concentrations in HEN buffer (10 mM Hepes, 0.25 mM EDTA, and 2.5 mM NaCl), held
at room temperature for 30 minutes, and then fixed with 0.1% glutaraldehyde for 4 hours at 4oC
before overnight dialysis into HEN. Samples were diluted, applied to glow-discharged carbon
coated grids, and positively stained with 1% aqueous uranyl acetate followed by extensive
washing. Grids were examined in a Tecnai 12 TEM operated at 100 KV in the tilted darkfield
mode, and digital images were recorded using a TVIPS 2024x2024 CCD camera.
Analysis of binding curves- Binding analyses typically refer to measurements of the
amount of bound complex formed as a function of increasing concentration of ligand. However,
for a variety of reasons, protein-DNA complexes often cannot be unequivocally identified and
quantified directly from gel retention assays. The free template concentration, m, is more easily
estimated and is directly related to the chemical potential of the DNA template in the equilibrium
solution (Georgel and Robert, 2002). A classical approach to the formulation of multiple binding
equilibria is through the binding polynomial or binding partition function, P, which is a function
of all relevant protein-binding reactions to the template (Wyman and Gill, 1990; Robert et al.,
1988). For a non-aggregating system, the chemical potential, µM, is related to the binding
polynomial by the relation
µM = µMo- RT ln P(x) (1),
56
in which µMo is the chemical potential under reference-state conditions (here defined in the
absence of ligand), and x is the Sir3p ligand concentration. The concentration of free-DNA
template, m, relative to the total concentration, mtot, is obtained by
m/mtot = 1/P(x) (2).
For a system in which concentrations are low enough for activity coefficient effects to be
neglected, P(x) is simply a sum of terms, each representing the concentration of a DNA species
(bound by zero or more Sir3p ligands) relative to the unbound DNA template concentration.
(The simplest example is a single-site binding model, or P = ([M] + [MX])/[M] = 1 + k x, where
k is the affinity constant in M-1.) Several binding polynomials (P) representing different binding
models were formulated following standard procedures (e.g., Robert et al., 1988; Teif, 2007) and
tested against the gel-retention assay data.
Independent sites: P1 = (1 + k x)n (3a),
Sites interacting in pairs: P2 = (1 + 2 k x + delta k2 x2)n/2 (3b),
Perfectly cooperative model: Pn = 1 + kn xn (3c),
Nearest-neighbor model: Pnn = (1,1) An (1, 0)' (3d).
In these equations k represents the intrinsic site affinity (units M-1), the dimensionless
parameter delta gives the multiplicative increase in affinity due to cooperative interaction with
another bound site, n represents the number of Sir3p binding sites on the template, and A is a 2x2
transfer matrix with the top row (1, 1) and the bottom row (k x, delta k x).
57
Given xtot and mtot, the total concentrations of Sir3p and DNA template, respectively, the
free concentration of Sir3 in the equilibrium solution was obtained by solving the mass-balance
equation: the total Sir3 concentration is written as the sum of the free and bound Sir3, the latter
being expressed in general form by a logarithmic derivative of the binding polynomial (Teif,
2007), giving
xtot = x + (d ln P(x)/ d ln x) mtot (4).
Eq. (4) implicitly defines the free Sir3p concentration, x, as a function of the total Sir3p
and DNA concentrations and the other binding parameters contained in P(x). A bisection
algorithm was used to obtain x at each total Sir3p concentration. Binding parameters were
determined by fitting the free DNA template fraction (eq. 2) versus total Sir3p concentration
using the Marquardt algorithm for the different models for P(x) given in eq. 3.
RESULTS
We have recently demonstrated that Sir3p can interact directly with long dsDNA (~2500
bp) and that the DNA component(s) of chromatin contributes to its recruitment (Georgel et al.,
2001; McBryant et al., 2008). As previous studies have strongly suggested that Sir3p does not
display any significant DNA sequence specificity (Shore and Nasmyth, 1987; Buchman et al.,
1988; Johnson et al., 1990), we focused our attention on binding as a function of DNA
conformation and size. To further characterize the contribution of the DNA component of
chromatin, we performed numerous EMSA and Electron Microscopy (EM) imaging experiments
58
to precisely assess the effect of DNA length and conformation, as well as Sir3p molarity on
binding and formation of higher-order complexes.
Binding ability of Sir3p to DNA and NAs.
As controls, the binding of Sir3p to 208-12 normal and tailless nucleosomal arrays (NA)
and parent 208-12 DNA was confirmed by EMSA under previously described conditions
(Georgel et al., 2001). The results indicated that Sir3p appears to bind cooperatively to the DNA
template (Figure 2.1). We also noted that the lack of histone tails does not appear to impede
Sir3p binding to nucleosomal arrays (compare Figure 2.1, panels B and C, lanes 11-14 and lanes
18-21). As the molar ratio of Sir3p was increased, we observed the formation of very large
DNA-Sir3p complexes, represented by material that did not migrate very far from the wells (see
white arrowhead in Figure 2.1, panel A, lane 7). Under this set of experimental conditions (50 ng
of DNA in a final volume of 10 µl), the phenomenon was observed only with the 208-12 DNA
template. This observation confirms the formation of the previously described Sir3p-associated
supra-molecular complexes (Georgel et al., 2001).
59
Figure 2.1: Sir3p binding efficiency to DNA is similar to that of tailless NA Panel A. Binding of Sir3p to 208-12 DNA. Lane 1 contains the 1kb+ DNA marker. Lane 2 corresponds to 208-12 DNA alone. Lanes 3 to 7 are binding of Sir3p to 208-12 DNA at increasing rSir3 ratios (from 1 to 16 Sir3 per 208 bp of DNA). The white arrowhead (lane 7) indicates the position of supra-molecular complexes. Panel B. Binding of Sir3p to 208-12 NA. As described above lane 8 contains the 1 kb+ DNA marker and lane 9 is NA alone. Lanes 10-14 correspond to NA in presence of increasing rSir3 ratios. Panel C. Binding of Sir3p to 208-12 tailless NA. Trypsinized arrays were used as templates. Lane 15: 1 kb+ DNA marker. Lane 16: NA tailless alone. Lanes 17-21 correspond to NA in presence of increasing rSir3 ratios. All EMSA were performed under high DNA concentration (10 ng/µl).
Length-dependence: Sir3p binds differently to DNA fragments shorter than 208 bp.
To test Sir3p binding as a function of DNA length in a context relevant to chromatin, we
first monitored Sir3 binding activity to fragments that could accommodate one, seven and twelve
nucleosomes. As we increased the size of the DNA from 208 bp using single unit or tandemly-
repeated copies of the 208 bp 5S rDNA from Lytechinus variegatus (Georgel et al., 1993)
(Figure 2.2, A, panel A) to 1456 bp (seven repeats of 208 bp DNA, Figure 2.2, A, panel B) and
2496 bp (twelve repeats of 208 bp DNA, Figure 2.2, A, panel C), we did not observe any
60
significant difference in binding behavior. Sir3p molar ratios (rSir3, normalized to 208 bp) from 1
to 16 were tested. At rSir3 from 4 to 16, the Sir3p-DNA complexes that forms are heterogeneous
as seen by the presence of a smear (see Figure 2.2, A panels A-C lanes 4-6). Based on the
determination of free-DNA half-depletion for each DNA length, no significant difference was
observed as a function of the DNA size, suggesting that the actual number of 5S rDNA repeats
does not play a critical role in Sir3p binding. Interestingly, as the molar ratio is increased, the
Sir3p-DNA complexes become too large to migrate significantly in the 1% agarose gels (as seen
by accumulation of material in the wells). In conclusion, the Sir3p affinity for DNA fragments in
the size range from 208 to 2496 bp did not appear to vary significantly.
Figure 2.2, A: Binding of Sir3p as a function of DNA length. Panel A: Binding of Sir3p to 208-1 DNA. Lane 1 corresponds to 208-1 DNA alone. Lanes 2-6 correspond to 208-1 DNA in presence of increasing rSir3 ratios (from 1 to 16 Sir3p per 208 bp of DNA). Location of DNA marker fragments indicated on the left side of each panel. Panel B: Binding of Sir3p to 208-7 DNA. Lane 1: 208-7 DNA fragment alone. Lanes 2 to 6, similarly to Panel A, are 208-7 DNA in presence of increasing rSir3 ratios (from 1 to 16). Panel C: Binding of Sir3p to 208-12 DNA. Lane 1: 208-12 DNA alone. Lanes 2-6: 208-7 DNA in presence of increasing rSir3 ratios (from 1 to 16).
61
Our initial results indicated that Sir3p binding to NAs leaves the linker DNA highly
accessible for restriction enzyme cleavage (Georgel et al., 2001), although more recent studies
have suggested a possible role for linker DNA in Sir3p binding (McBryant et al., 2008). To
further investigate this issue and to determine the minimal length requirement for Sir3p binding,
we performed additional EMSA experiments with DNA fragments ranging from 12 to 100 bp
(Figure 2.2, B). Note that agarose gels did not give a sufficient resolution and that we switched to
an acrylamide system to investigate the formation of Sir3p-DNA complexes. Sir3p was added to
the DNA at rSir3 ranging from 1 to 8. In all cases, Sir3p binding was very efficient, even at rSir3 of
1 to 2 (Figure 2.2, B, lanes 3 and 4 for each panel). Higher molecular weight complexes were
observed (as indicated by accumulation of material close to the wells, see open triangles in
Figure 2.2, B) for all DNA templates. Well-defined complexes were observed with DNA
fragments of 32 bp and 12 bp at rSir3 of 1 to 4. When the rSir3 was increased to 8, additional bands
(Figure 2.2, B, black triangles) corresponding to higher molecular weights were present,
suggesting oligomerization of the initial Sir3p-DNA complexes. Combining this observation
with earlier results using 208-12 DNA or NA (Georgel et al., 2001) leads to a model describing
large supra-molecular complexes involving trans- interactions between individual DNA or NA
molecules. These observations are consistent with a role for short linear stretches of linker DNA
in Sir3p-chromatin fiber interactions.
62
Figure 2.2, B: Binding of Sir3p to short DNA fragments. Side by side comparison of Sir3p binding ability to DNA of size ranging from 100 bp down to 12 bp. The DNA and Sir3p-DNA complexes were separated using a native 6% acrylamide gel. Lane 1 in each panel contains the 1 kb+ DNA marker. Lanes 2 correspond to DNA alone. Lanes 3-6 contain DNA in presence of increasing rSir3 ratios (from 1 to 8). The white arrowhead indicates the location of supramolecular complexes (close to the well). The black arrowheads indicate the formation of an intermediate size complex (32-bp DNA and 12-bp DNA panels). The sizes of relevant DNA bands are indicated on the left of each panel.
DNA sequence/conformation dependence on Sir3p binding.
Our experiments indicate that both nucleosomal and DNA components may be involved
in Sir3p-chromatin interactions (McBryant et al., 2006; see above). We were therefore interested
in whether Sir3p showed a preference for DNAs with different conformations. Specifically,
linker DNA may have a more linear conformation, where nucleosomal DNA would display a
more curved or bent conformation. Based on the nucleosome core particle structure (Luger et al.,
1997), nucleosomal DNA is distorted at positions 45 and 100 (positions nearly opposite the
nucleosomal dyad axis), showing a more pronounced bending. The rest of the nucleosomal DNA
is more gently curved around the histone octamer. To mimic all possible DNA conformations
described for the nucleosome core particle, we tested three different DNA fragments that would
63
represent 1) the linker DNA as a linear/straight fragment (221 bp covering position -185 to +36
from the hsp70 Drosophila melanogaster promoter region, which displays a nearly straight
conformation), hereafter referred to as “Straight”, 2) the more pronounced bends described over
the nucleosomal DNA at positions 45 and 100 were mimicked using a template containing a
sharper bend (a 199 bp fragment covering the binding site for RNA polymerase 1 from
Acanthamoeba histolitica (Georgel and Robert, 2002), referred to as “Bent”, and 3) the more
evenly curved nucleosomal DNA, using the known 5S rDNA positioning sequence that contains
two curves (Georgel et al., 2001, Georgel and Robert, 2002), referred to as “Curved”. The actual
conformation of all three fragments was modeled using an algorithm developed by Trifonov
(Trifonov, 1991) (http://hydra.icgeb.trieste.it/~kristian/dna/, Figure 2.3, A). In Figure 2.3, A, all
fragments are represented in an orientation that reflects the maximum predicted curvature. To
confirm the predicted DNA conformation differences, we performed QAGE analysis (in
triplicate) of the various fragments to determine their electrophoretic effective radii. If the
conformation of all fragments was assumed to be linear, then based on size only, the order of Re
would have been: “Straight”, “Curved”, followed by “Bent”. However, our results (see Table
2.1) indicate a conformational component that confirms the predictions from the Trifonov
modeling algorithm for the various DNA fragments. The smallest Re was that of the “Curved”
fragment (9.7 +/-0.3 nm), and the largest was that of the “Straight” DNA (11.18 +/- 0.6 nm). The
“Bent” DNA fragment had a value between those of the other two fragments (10. 37 +/- 0.46
nm). All three fragments were subjected to EMSA analysis under the lower concentration (5
ng/µl) conditions described for Figures 2.1 and 2.2. EMSA experiments employed a fixed mass
of DNA (50 ng) that was incubated in the presence of increasing rSir3. Comparison of half-
depletion of free DNA in each panel (Figure 2.3, B, panels A-C, lanes 7 and 8; rSir3 of 8 to 16,
64
Figure 2.3, C) showed the “Straight” fragment to be the template with highest Sir3p binding
efficiency and the “Curved” fragment (5S rDNA 208-1) to have the lowest binding efficiency.
The difference in binding EMSA shifts between the three fragments was very small. To confirm
this result, we performed a competition assay mixing all three fragments in equimolar amounts
with increasing rSir3. The order of shifting as seen by EMSA matched the expected binding
ability seen when using individual fragments (data not shown).
65
Figure 2.3, A and B: Predicted curvature and EMSA analysis of Sir3p binding to various DNA conformations
(A) The DNA sequences of all three fragments (p29-p58), PX199 and 208-1 were analyzed and visualized using the Trifonov algorithm (http:hydra.icgeb.trieste.it/~kristian/dna/). The presented orientations were selected to display the maximum curvature. The locations of the bends and curves are indicated by the white arrows. (B) Panel A: Sir3p-mediated DNA depletion analysis using the “Straight” DNA template. Lane 1 contains the 1kb+ DNA marker. Lane 2 contains the DNA template alone. Lanes 3 to 8 are the DNA templates in presence of increasing rSir3 (from 1 to 16). Lanes 8 to 11 contain a serial dilution of “Straight” DNA used for calibration in the depletion analysis. Panel B: Similar experiment using the “Bent” DNA as template. Panel C: Similar experiment using the “Curved “ DNA as template. The sizes of relevant DNA bands are indicated on the left of each panel.
The binding of Sir3p to short DNA stretches left open the possibility that the protein
could interact with the nucleosomal entry-exit DNA. A four-way-junction (FWJ) DNA template,
commonly used as a substitute for the nucleosomal DNA entry-exit region, was assembled and
used as a template (Panyutin and Hsieh, 1993). The binding affinity of Sir3p for the FWJ, based
66
on half-depletion of free DNA, was significantly lower than that of any of the tested linear
fragments (compare Figure 2.3, B, panel D, lanes 6 and 7 to Figure 2.3, B, panels A-C, lanes 6
and 7). This result suggests that Sir3p’s ability to bind to the nucleosome entry-exit region is
low. It is important to note that, depending on Mg+2 and EDTA concentrations, FWJ can adopt a
partially flattened conformation that may interfere with Sir3p binding (Lilley, 2000; Declais et
al., 2003). This potential conformational change may have affected Sir3p binding resulting in the
observed low affinity. All three linear DNA fragments' binding properties were tested by EMSA
and modeled, the results are summarized in Table 2.1.
Sir3p binding to DNA is highly cooperative.
The analysis of Sir3p-DNA binding indicates that it is strongly cooperative. To determine
the mode of interaction over a wide range of rSir3, the 208-1 monomeric unit of the tandemly-
repeated 208-7 and 208-12 DNA was selected to establish a cooperative binding baseline for
Sir3p. It was mixed in the presence of increasing rSir3, and free DNA depletion was then
determined from gels as a function of the total Sir3p concentration (expressed in µM) and
analyzed using eq. 2. The data, run in triplicate, provided a measure of the intrinsic run-to-run
variability of 0.1-0.15 in units of fractional template concentration. The tendency of the free
DNA to disappear completely after a few steps of the titration with Sir3p was clearly seen in
each experiment.
In order to determine how cooperatively Sir3 was binding, we tested several different
binding models using free-DNA depletion data for the 208-1 template and for the “Straight” and
“Bent” DNA templates. We first applied a non-cooperative independent-sites model (eq. 3a) with
a single class of binding sites. The results of this binding model were very poor: the standard
error of a point of the best fit varied from 0.20 to 0.27 depending on the template, and the non-
67
cooperative model was clearly unable to account for the steepness of the free-template depletion
curve. It bears pointing out that inclusion of additional classes of independent binding sites
would have exacerbated the poor fit, leading to apparent anti-cooperativity and consequently an
even less-steep curve (Wyman and Gill, 1990). Adding a cooperative interaction for pairs of
interacting Sir3 binding sites (eq. 3b) improved the fit somewhat, but the standard error of a
point of the resulting best-fit (0.14 - 0.23) was still far larger than the intrinsic errors associated
with the data.
We found that models best representing the free-DNA depletion data involved a high
degree of cooperativity in Sir3p binding to the template. The high cooperativity was first
modeled as a phenomenological "all-or-nothing", or perfectly cooperative, association (eq. 3c).
This simple model accounted well for the steepness of the template depletion curves and
provided standard errors of 0.1 - 0.15, well within in the range of the intrinsic run-to-run
variability of the data. Another cooperative model that has the additional benefit of being
immediately applicable to template DNA of any length is the nearest-neighbor model (eq. 3d).
This model described the data with the same standard error of a point as the perfect cooperativity
model. Although gel data are not of sufficiently high precision to distinguish between details of
the cooperative model chosen to represent the Sir3p interaction, the binding results clearly
suggest a high degree of cooperativity in their binding to the naked DNA template. The fits to the
data are shown in Figure 2.3, C. Intrinsic association constants obtained for Sir3p binding to the
different templates (“Straight”, Bent”, and “Curved”) using this model are shown in Table 2.1.
All templates show a similar number of binding sites for Sir3. Binding to the straight DNA is
favored by a factor of 10 over the curved template, although the cooperative interaction is
highest for the curved DNA.
68
Figure 2.3, C: Fraction of free DNA for increasing concentrations of Sir3 protein Left: "Straight" template (Hsp), Middle: "Bent" template (Pol1), right: "Curved" template (208-1). Estimated total Sir3
concentrations at 1/2 DNA depletion are 0.24, 0.28, 0.40 µM, respectively.
We also used EMSA to examine Sir3 binding to DNA templates containing 7 or 12
repeats of the 208 bp sequence. The free-DNA depletion data for the 208-1, 208-7, and 208-12
templates are shown in Figure 2.4. As with the 208-1 data alone, the 208-7 and 208-12 template
data are clearly consistent with cooperative Sir3p binding: a non-cooperative binding model
again resulted in a very poor fit. In order to analyze this data obtained for templates of
significantly different lengths, the nearest-neighbor interaction model was used (eq 3d). This
formulation is more physical than the perfect-cooperativity approach and models the
cooperativity as arising from stabilizing interactions between Sir3p molecules bound at
neighboring sites on the DNA template. This model also described the cooperative Sir3p binding
quite well: the curves shown in Figure 2.4 correspond to the best fit of this model to all datasets
simultaneously, in which the total number of sites was defined to be 1, 7, or 12 times the 208-1
69
template value. Fitted parameters are given in the legend of the figure. To illustrate the positive
cooperativity in Sir3p binding, the first line of Table 2.1 shows the intrinsic binding constants,
presented as dissociation constants. These constants, with values on the order of 100 µM, refer to
binding of a Sir3 protein to a site on the DNA when no Sir3 is bound to adjacent sites. For
contrast, the third line of Table 2.1 shows the effective binding constant for Sir3p at a site
neighboring a bound Sir3 protein. These values are on the order of 0.1 µM. The affinity increase,
given by the parameter delta, is a few hundred to a thousand-folds for all 208-bp repeat
templates.
Figure 2.4: Sir3p binds cooperatively to DNA. Fraction free/total DNA template plotted as a function of the logarithm of the total Sir3p concentration for the 208-1 (open circles), 208-7 (filled circles), and 208-12 (white squares) templates. Corresponding lines (solid, broken, and dotted, respectively) indicate the best-fit of the nearest-neighbor model (eq. 3d, best-fit SEP 0.10) with binding site size of 28±2 bp (calculated by template length in bp / n), intrinsic binding constant log(k/1µM-1)=-1.88±0.03 for the 208-1 template and -2.09±0.02 for the 7 and 12-mers, and nearest-neighbor interaction term log(delta)=2.47±0.04, 2.66±0.03, and 3.22±0.08 for the 208-1, -7, and -12 templates, respectively.
70
The formation of 208-12 DNA-Sir3 complexes and cooperativity were further
investigated by EM imaging (Figure 2.5, panel D) to evaluate stoichiometry and conformation.
Cooperative binding seemed to be favored under the high DNA and high Sir3p concentration
conditions, suggesting a Sir3p-Sir3p interaction component driven by high local concentration.
The different complexes observed by EMSA were imaged at various rSir3. At low rSir3 and low
DNA concentration, cooperative binding is evidenced by the co-existence in the same sample of
free DNA fragments (Figure 2.5D, panel A) and Sir3p clusters on the DNA (panels C, D). As the
rSir3 is increased to 8 and 16 and under high DNA concentration conditions, the DNA becomes
coated with Sir3 clusters (Figure 2.5, panels E and F) and eventually forms self-associating
DNA-Sir3p complexes (Figure 2.5, panels G and H). Once again the cooperative nature of the
event is evidenced by the presence of Sir3-free DNA molecules adjacent to Sir3p-associated
DNA complexes (Figure 2.5, panels D and G).
Table 2.1: Characterization of mobility and Sir3 binding of the DNA fragments
71
Figure 2.5: EM imaging of Sir3p-DNA complexes at Low and High DNA concentrations. 208-12 DNA at a concentration of 5 ng/µl and rSir3=4 (panels A-C). Small (black arrow) and large oligomers of Sir3p (grey arrow) can be observed in the background in the presence of free DNA (panel A). The population also includes complexes with large Sir3p clusters bound DNA (white arrow) and resulting in loops (panels B and C). The bar in Panel C corresponds to 20 nm. At higher rSir3 of 8 and high DNA concentration (panels D-F), free DNA can still be observed (panel D), in addition to molecules that are almost entirely cover by Si3p (panels E and F). Free Sir3p can still be observed (black arrows). At even higher rSir3=16, the Sir3p-coated DNA complexes appear to associate to form very large supra-molecular assemblages (white arrows, panels G and H). The co-existence of free DNA alongside the large assemblages (panel G) strongly supports the cooperative model.
Sir3p saturation affects complex formation.
Based on our observation linking the formation of higher order Sir3p-DNA complexes
with the actual DNA concentration, we decided to systematically evaluate its effect on the initial
Sir3p binding, as well as on the formation of supra-molecular complexes. EMSA experiments
were performed using the same three previously described DNA, NA and tailless NA templates
under higher DNA concentrations over the same range of Sir3p molar ratios. The relative
binding affinity observed under 5 ng/µl DNA concentration conditions reflected the initial results
72
obtained and described in Figures 2.1, C and 4, A-C. The connection between the formation of
supra-molecular complexes through DNA and Sir3p interactions was investigated using 10 ng/µl
DNA (Figure 2.6 A, lanes 13-14, see white arrowhead). No large complexes (running close to
wells in agarose gels) were observed under 10 ng/µl DNA conditions with either the NA or
tailless NA (Figure 2.6A, B, and C, lanes 6-7 and 13-14). Further, we observed the formation of
a well-defined complex and the depletion of free template at a lower rSir3p under 10 ng/µl DNA
experimental conditions for NA, tailless NA, and naked DNA (Figures 2.6, A, B, and C:
compare lanes 6 and 7 to lanes 13 and 14). The later depletion of free material and the early
formation of complexes of intermediate size between Sir3p and tailless NA suggest a role for the
histone N-termini, possibly through electrostatic interactions resulting in a partial masking of the
DNA. The results indicated a strong connection between local Sir3p concentration and DNA
binding efficiency. This suggests that the local Sir3p concentration may contribute to the
formation of complexes involved in silencing in a DNA-dependent manner.
73
Figure 2.6: Effect of DNA concentration on Sir3p binding. (A) Effect of DNA concentration alone on Sir3p binding. Side by side comparison of Sir3p binding ability at 5 ng/µl (left panel) and at 10 ng/µl (right panel). Lanes 1 and 8 contain the 1 kb+ DNA marker. Lanes 2 and 9 correspond to DNA alone. Lanes 3-7 and 10-14 contain 208-12 DNA in presence of increasing rSir3 ratios (from 1 to 16). The white arrowhead indicates the location of supramolecular complexes (7). (B) Effect of DNA concentration on Sir3p binding to nucleosomal arrays. The same experiment was repeated using NA instead of DNA. The gel set-up is identical to that described in Figure 2.6 A. (C) Effect of DNA concentration on Sir3p binding to tailless nucleosomal arrays. The same experiment was repeated using tailless NA instead of DNA. The gel set-up is identical to that described in Figure 2.6 A.
74
Sir3p interaction with single-stranded DNA
After observing that Sir3p can bind to short dsDNA fragments, we investigated whether it
can also complex with single-stranded DNA (ssDNA). The primer p7 (32mer) was used as
template for Sir3p EMSA analysis (Figure 2.7). The 32mer ssDNA complexed with Sir3p even
at a ratio of one Sir3p molecule per DNA molecule, and was nearly entirely depleted at a rSir3 of
4 (Figure 2.7, lane 5). Multiple complexes of higher molecular weight were formed as the rSir3
was increased, indicating either multiple binding sites or the formation of Sir3-DNA oligomers.
At rSir3 of 4 and 8 (Figure 2.7, lanes 5 and 6), the complexes generated were too large to be
resolved in the gel, and some of the material remained trapped next to the well area.
Figure 2.7: Sir3p binds to ssDNA. EMSA experiments were performed using 32mer ssDNA (primer p7, see sequence in material and methods), under similar DNA concentration and rSir3 conditions to that shown in Figure 2.2 B. Based on DNA depletion, Sir3p appears to bind short ssDNA as efficiently as short dsDNA forming complexes of various sizes (white arrows). Lane 1 contains 1 kb+ DNA marker. Lane 2 contains ssDNA alone. Lanes 3 to 6 are ssDNA in presence of increasing rSir3 (1 to 8).
75
DISCUSSION
The ability of Sir3 to contribute to the general silencing mechanism is well-established
(Loo and Rine, 1994; Hecht et al., 1995; Hecht et al., 1996). The development of silenced
regions over the HM loci and telomeres has been linked to Sir3p interactions with hypoacetylated
histone H3 and H4 N-termini (Rine and Herskowitz, 1987; Loo and Rine, 1994; Hecht et al.,
1995), but despite initial studies (Moazed et al., 1997; Liaw and Lustig, 2006) the mechanism of
spreading along chromatin fibers remains poorly explained. We have shown that the Sir3p
interacts with both DNA and the non-tail nucleosomal components of chromatin while
establishing condensed chromatin architecture. Our results presented here, as well as previously
published work strongly suggest that the DNA component may be important in the establishment
of silenced regions. Sir3p-DNA interactions have previously been mentioned in the literature
(Georgel et al., 2001; McBryant et al., 2008), but no thorough analysis had been performed to
investigate their actual contribution to silencing. The description of the DNA binding properties
of Sir3p as cooperative, size-, conformation-, and DNA concentration-dependent opens the door
for a more a thorough understanding of the how Sir3p functions in the initiation and spreading
of silencing.
Importance of nucleosomal DNA features for Sir3p binding.
Our Sir3p binding analysis indicates high affinity for naked DNA, cooperative binding at
higher rSir3p, and importantly, the influence of local Sir3p concentration, all of which support a
model where initial Sir3p recruitment and spreading would consist of three separate steps. The
initial recruitment of Sir3p to a chromatin or DNA template in vivo has been shown to involve
Sir4p (Hecht et al., 1995) and histone tail deacetylation by Sir2p (Tanny et al., 1999), brought
together through interactions with the H3 and H4 N-termini. The level of acetylation of histone
76
H3 appears to be intricately linked to the regulation of the efficiency of the initial Sir3p
mobilization and may contribute to its ability to spread (Kristjuhan et al., 2003). However,
Hoppe and co-workers (Hoppe et al., 2002) recently presented evidence for an alternative mode
of action that would involve Sir3p in spreading silencing in the absence of Sir2-Sir4 complexes
(see also Hecht et al., 1996; Renauld et al., 1993; Strahl-Bolsinger et al., 1997). Following this
line of evidence, one can envision a significant role for DNA in spreading of silencing that may
not be dependent on histones H3 and H4.
Our observation that the binding of Sir3p to 208-12 5S rDNA displays a similar affinity
to that of the same DNA reconstituted into nucleosomal arrays using tailless histones confirms
that Sir3p plays a role in spreading silencing through DNA interactions (Georgel et al., 2001,
Figure 2.1). Our analysis of DNA size dependence seems to indicate that DNA larger than 208
bp is a less efficient target than shorter fragments (down to 12 bp). This may include association
with short linker DNA, if available for binding. This observation, combined with the ability of
Sir3p to spread silencing in the absence of Sir4p and Sir2p, suggests a spreading mechanism
involving short stretches of nucleosomal or linker DNA that would require minimal Sir3p
interactions with core histones. Our analysis of DNA conformation indicates a Sir3p binding
preference for short straight DNA fragments. At equal molar ratio, potential interactions between
Sir3p and nucleosomal DNA could be mediated through recognition of short stretches of straight
DNA. As was described by Luger et al. (Luger et al., 1997), the last 10-bp segment at the edge of
the nucleosome structure (close to the entry-exit location) is mostly straight and has limited
interactions with the histone octamers, and thus could provide Sir3 with a preferred DNA
binding site. Another physical feature of the nucleosome core particle may also be of importance
for Sir3p binding to DNA. The periodicity of histone tails passing through the DNA double helix
77
at 20 bp intervals (Luger et al., 1997) may delineate structural domains that could be preferred
targets for Sir3p binding. The 20 bp interval is of the same order of magnitude as the minimal
observed size that allows Sir3p binding (12 bp, as described in our size-dependence binding
analyses, Figure 2.2, B). The observed differences in binding efficiency to linear DNA of ~ 200
bp in length are also consistent with such a mode of interaction. The mostly straight DNA was a
slightly better substrate for Sir3p binding, whereas the DNA with highest curvature interacted
less efficiently. In the context of nucleosomal DNA, the highest curvature is observed at super-
helical (SHL) positions 1.5 (15 bp from the DNA entry site) and 3.5 and 4 about 15 bp from
straighter DNA (around SHL position 5.5) having fewer histone contacts (Luger et al., 1997;
Pryciak and Varmus, 1992; Muthurajan et al., 2003). The DNA fragments we tested (Figure 2.3,
panels A-C) reflect such differences in intrinsic or induced curvature. Our results suggest that
Sir3p binding may be more efficient over straighter portions of the nucleosomal DNA. Since
other proteins such as MENT (Grigoryev and Woodcock, 1998) have been described as cross-
linking nucleosomal DNA at the entry-exit region, we investigated Sir3p binding to a synthetic
FWJ designed to mimic the nucleosomal linker entry-exit site. The poor binding of Sir3p to such
templates does not support a binding mechanism similar to that of MENT or even linker histones
(Figure 2.3, panel D; Grigoryev and Woodcock, 1998; Thomas et al., 2002), but instead favors a
model involving association with nucleosomal and/or linker DNA.
78
Figure 2.8: Model for DNA-mediated Sir3p binding component. The model represents only the DNA portion of a chromatin fiber (the linker DNA is highlighted in red and the histone N-termini in green). The blue ovals represent Sir3p. The initial Sir3p recruiting step is mediated by nucleosomal domain(s) interactions. As the local concentration increases (indicated by the blue triangle), additional Sir3p-DNA interactions stabilize the complex. Upon further increase in local concentration, Sir3p cooperative binding to DNA significantly contributes to the spreading of silencing along the chromatin fibers (as indicated by the diverging blue arrows).
Cooperative binding and Sir3p local concentration.
Our results strongly suggest a cooperative binding mechanism for Sir3p to both DNA
and NA (Figures 2.4 and 2.5). The binding mechanisms that best described the EMSA-based
data is the nearest neighbor model (see Table 2.1). The presence in EM images of DNA entirely
devoid of Sir3 proteins adjacent to templates with high Sir3 occupancy (Figure 2.5, D) at rSir3
ratios as high as 16 also strongly supports our EMSA analysis. Similar gel results for Sir3p have
recently been reported (McBryant et al., 2008). In addition to the observed cooperativity of Sir3p
binding to DNA and NA, we observed that at identical rSir3 ratios, the formation of Sir3p-DNA or
79
Sir3p-NA complexes was highly dependent on the actual substrate molarity at which the
experiments were performed (Figure 2.5, A-C). Increases in local Sir3p concentration appear to
favor cooperative recruitment of Sir3 molecules to the DNA and chromatin fibers, indicative of a
novel mode of interactions leading to silencing. The formation of higher molecular weight (seen
close to the well during EMSA experiments, Figure 2.5, A, lanes 13 and 14, and (Georgel et al.,
2001)), referred to as supra-molecular complexes, was also shown to be highly dependent on
molarity. The high rSir3 ratios combined with high molarity leads to the coating of DNA
molecules, and also contributes to the formation of assemblages of Sir3p-covered DNA (see
Figure 2.5, D, rSir3 =16, at high molarity conditions). The combination of results obtained with
DNA and tailless NA suggests a limited role for the histone tails in Sir3p spreading. The
cooperative nature of Sir3p binding and the importance of local concentrations lead us toward a
model where Sir3p silencing would be described as a three-step mechanism (Figure 2.8) that
bears some similarities to the binding and recruitment of heterochromatin proteins 1 or HP1 , ,
and , each playing a complementary role (Vershure et al., 2005; Hediger and Gasser, 2006). The
initial recruitment of Sir3p would be dependent on interactions with nucleosomal features that
involve binding to structurally-specific DNA conformation features of the nucleosome core
particle (see previous section). As the local concentration of Sir3p reaches a critical value, the
cooperative mode of binding would favor Sir3p spreading from occupied sites using DNA as a
track. The binding of Sir3p to DNA also appears to initially promote the formation of DNA
loops (see Figure 2.5, D, panels B and C) that are reminiscent of the MENT-induced compaction
of chromatin (Grigoryev, 2001). This model is supported by evidence that Sir3p alone, when
over-expressed in yeast, can account for repression observed at specific loci (Hecht et al., 1996;
Hoppe et al., 2002; Renauld et al., 1993). As the local concentration reaches its threshold, the
80
Sir3p-coated DNA fiber would start to interact, probably through DNA-Sir3/Sir3-DNA bridges.
The actual role of the tails of histones H3 and H4 and their post-translational modifications
remains to be investigated. To further delineate the Sir3p domains responsible for spreading of
silencing, we have also started a more thorough analysis of Sir-3 DNA affinity using various
Sir3p deletion constructs.
Sir3p binding to single-stranded DNA.
The binding of Sir3p to ssDNA may contribute to the establishment of functional
telomeric complexes. Telomere silencing and telomere position effect (TPE) are both associated
with the presence of Sir3p in sub-telomeric DNA, but are also regulated by the amount of
ssDNA available (Bourns et al., 1998). A specific class of single-stranded binding proteins,
including Est1 (Virta-Pearlman et al., 1996), NSR1 (Lin and Zakian, 1994), GBP2 (Lin and
Zakian, 1994), and the yeast protein Rlf6p and Chlamydomonas reinhardtii Gbp1 (Konkel et al.,
1995), has been described as binding single-stranded G-strand telomeric DNA, and it remains
possible that Sir3p also plays a role in this process through its ssDNA binding activity. This
mechanism may have similarities to that observed using truncated forms of Rap1 (Konkel et al.,
1995; Kyrion et al., 1992). Konkel and colleagues have shown that Rlfp6p is required for the
appropriate location of Rap1p, a mechanism that also involves Sir3p and Sir4p (Rine and
Herskowitz, 1987; Kyrion et al., 1992). A possible role for Sir3p, consistent with its ssDNA-
binding activity, may include the stabilization of the Rlf6p-Rap1 complex (de Bruin et al., 2000).
Alternatively, the ss-nucleic acid binding activity of Sir3p may be functionality related to that of
DDP1, a heterochromatin-associated in Drosophila involved in heterochromatin maintenance
and silencing (Huertas et al., 2004). Further investigations will be necessary to precisely
81
determine the nature of the role played by ssDNA binding at the yeast mating-type loci and
telomeres.
ACKNOWLEDGMENTS: We thank Dr. Lustig for providing the pJC52 plasmid, Drs. Hager and John
for critical reading of the manuscript.
82
CHAPTER 3: DIFFERENTIAL BINDING OF WILD TYPE AND RTT MeCP2 MUTANTS TO HISTONE TAILS MEDIATES CHROMATIN
INTERACTIONS.
Nicholas L. Adkins1, Steve J. McBryant2, Nina Akimenko4, Jennifer M. Leidy1,3, Christopher L.
Woodcock4, Jeffrey C. Hansen2 and Philippe T. Georgel1
AUTHOR ADDRESS: Department of Biological Sciences and Cell Differentiation and
Development Center, Marshall University1, West Virginia 25701, Department of Biochemistry
and Molecular Biology, Colorado State University2, Fort Collins, CO 80523-1870, School of
Medicine, West Virginia University3, Morgantown, WV 26506-9100, Paris Sud, Orsay 91405
France, Department of Biology, University of Massachusetts, Amherst4, Amherst, MA 01003
AUTHOR EMAIL ADDRESS: [email protected]
CORRESPONDING AUTHOR FOOTNOTE: Address correspondence to: Philippe T. Georgel, PhD,
Marshall University, Department of Biological Sciences, 1 John Marshall Drive, Huntington WV 25755.
Fax; 304-696-7136.
To be submitted
83
ABSTRACT
Rett Syndrome (RTT) is a debilitating neurodevelopmental disorder that has been linked
to various mutations in the MECP2 gene in clinically diagnosed patients. The function of this
chromatin associated protein and the biological role of these mutations remain ill-defined. Initial
characterization described MeCP2 as a repressor of transcription that selectively binds to
methylated CpG dinucleotides. More recently, it has been described as a strong contributor to
chromatin condensation into higher order structures. In addition, this MeCP2-induced chromatin
condensation was found to be independent of methyl-DNA binding. The specific molecular
determinants for chromatin binding and compaction capability have remained mostly unknown.
Here, we demonstrate that MeCP2 interacts with histone N-termini selectively recognizing
histone H2A, H3 and H4 N-termini (tails). Acetylation of these tails does not significantly affect
binding to chromatin, but decreased MeCP2 ability to fold nucleosomal arrays. In contrast to the
wild-type MeCP2, we found that RTT-associated point mutant R133C recognizes all four histone
N-termini, suggesting a role for the histone H2B tail in the RTT phenotype. We analyzed another
RTT mutant, the R168X truncation, for histone tail interactions. This RTT mutant was
previously found to retain the capability of binding to chromatin, but not supporting the
formation of supramolecular complexes. This truncation was found to be associated with a loss
of interactions with histone N-termini. The observed binding pattern supports the hypothesis that
interactions with histone N-termini are required for proper MeCP2-driven chromatin
compaction. From these results, we concluded that histone tails are required for proper MeCP2-
chromatin interactions, and contributes to the folding of chromatin into higher order structures.
84
INTRODUCTION
The field of gene regulation has been increasingly intertwined with investigations of
epigenetic events. DNA methylation and the recruitment of specific proteins that can recognize
methylated cytosine bases (in the context of CpGs) is probably the best-characterized epigenetic
modification linking the transition between gene activation and repression (Cedar and Bergman,
2009). Histone post-translational modifications (PTMs) have also generated very strong interest
from researchers who focus on the regulation of gene expression (for recent review, see Berger,
2007). The identification of specific histone N-termini (or tails) PTMs has led to the creation of
an entire sub-field of research driven by the determination and characterization of the “histone
code”. In addition to DNA and histone chemical modifications, the chromatin contribution to
DNA-associated regulatory events can be linked with proteins that recognize either methylated
DNA or modified histone tails. These proteins, referred to as chromatin-associated proteins
(CAPs), can act as co-activators or co-repressors. They may also contribute to the chromatin
compaction/de-compaction equilibrium required for the accessibility of DNA regulatory
elements. The transition from a fully extended chromatin conformation, referred to as the 10nm-
fiber or “beads-on-the-string” to a more compacted 30nm-fiber, or even higher-order structure,
has long been shown to strongly influence transcription efficiency. Most of the research on
higher order chromatin structure has been performed using chromatin model systems consisting
of defined arrays of nucleosomes (nucleosomal arrays or NA) (reviewed in Luger and Hansen,
2005). The inherent ability of chromatin to fold under specific ionic strength conditions or
changes in Mg concentration can be stabilized and/or enhanced by the recruitment of certain
CAPs. These proteins can further enhance chromatin compaction and play a role in the
maintenance of transcriptionally repressed state. Among these CAPs, the methyl CpG binding
85
protein 2 (MeCP2) has been showed to be critical for proper neural development. Specific
MeCP2 mutations have been linked to the Rett syndrome (RTT), a progressive childhood
neurodevelopmental disorder that is one of the most common causes of mental disabilities in
female patients (Amir et al., 1999). MeCP2 has been initially described as a methyl DNA-
binding protein (Byrd), but can also play a significant role in chromatin compaction (Georgel et
al., 2003). Recently, MeCP2 has been shown to not only associate with repressed chromatin
regions, but also actively transcribed loci (Yasui et al, 2007, Chahrour et al., 2008). Regulatory
events associated with transcription regulation and chromatin folding may require the generation
of small domains of compacted chromatin that would affect chromatin tertiary structure that are
still compatible with active transcription (Georgel et al., 2003). These inconsistencies in MeCP2
behavior may indicate an additional level of signaling required to explain MeCP2s localization
and/or function.
The N-terminal tails of core histones serve as binding determinants for some CAPs with
particular arrangements of PTMs serving as signaling mechanisms for their function and
localization. The role of MeCP2 in repression has classically been described to occur through its
binding to methylated DNA and acting as a global transcriptional repressor (Lewis et al., 1992;
Nan et al., 1997). Previous studies using nucleosomal arrays as model chromatin demonstrate
that MeCP2 can condense chromatin fibers and even form oligomeric structures in the absence of
DNA methylation, suggesting that the ability of MeCP2 to silence gene expression may be
related to large-scale chromatin organization (Georgel et al., 2003). The concept of genomic
silencing mediated by MeCP2 is reinforced by its ability to recruit Sin3p and Histone De-
acetylase 1 (HDAC) to specific loci resulting in hypoacetylation of local histones (El-Osta et al.,
2002; Suzuki et al., 2003). The presence of HDAC suggests that histone N-termini acetylation
86
could contribute to MeCP2’s ability to bind and/or condense chromatin. Recently published data
demonstrates that MeCP2, when complexed with mono-nucleosomes, resides on nucleosomes in
close proximity to histone H3 by evidence of biotin transfer after protein crosslinking (Nikitina
et al., 2007). As a direct consequence of this observed proximity it has been assumed that H3
tails could be a determinant for MeCP2 recruitment. However, when MeCP2-mononucleosome
binding was analyzed, it appeared that the H3 tail was not required for MeCP2 recruitment
(Ishibashi et al., 2008). To date, no unequivocal information on role of other histone tails on
MeCP2 binding or function has been gathered.
In this study, we have investigated the importance of each individual histone N-terminal
tail on MeCP2 binding. We also monitored MeCP2-induced chromatin compaction and found
that the two functions of binding and compaction can be separated. Acetylation of histone tails
was previously demonstrated to have very little effect on MeCP2 binding to mono-nucleosomes
(Ishibashi et al., 2008). Our research confirms that MeCP2 binding to NA behaves similarly, no
drastic change in MeCP2 binding with addition of acetylation to the N-terminal tails; however,
we observed significant differences in MeCP2-induced compaction of NA when monitored by
electron microscopy (EM). Investigation on the role of individual tails on MeCP2 recruitment
indicates that the H2B tail is not a structural feature recognized by MeCP2. Histone N-termini
from H2A, H3, and to a lesser extent, H4 participate in MeCP2 binding. Interestingly, removal of
H3 tails affected recruitment of MeCP2 to arrays of nucleosomes in contrast to mono-
nucleosome binding studies (Ishibashi et al., 2008). Additionally, we investigated the effect of
two MeCP2 RTT mutants on binding to nucleosomal N-termini. The MeCP2 RTT point mutant
R133C has previously been shown to have a decreased binding affinity for methylated DNA
when compared to the wild type protein (Ghosh et al., 2008). Our binding assays also show in
87
vitro interaction of this R133C point mutant with all histone N-termini, in contrast to the wild-
type MeCP2 protein which only complexes with tails of H2A, H3, and, less efficiently, H4. The
RTT deletion mutant R168X which lacks the Transcription Repression Domain (TRD, see
Figure 3.1), previously shown to retain binding capability to NA, but lacks ability to form higher
order chromatin complexes. Our results indicate that the R168X fails to recognize histone N-
termini, possibly linking this loss of function to its inability to support formation of chromatin
higher order complexes. The residual binding of this truncated mutant may be DNA-driven or
could require other nucleosomal features. These in vitro binding studies directly demonstrate
MeCP2s capability to interact with core histone N-terminal tails and signify a potential novel
mechanism for MeCP2 localization and/or function. Furthermore, we describe differential
chromatin binding that occurs with two known RTT mutant MeCP2 proteins, possibly
identifying a molecular basis for RTT occurrence.
Figure 3.1: MeCP2 Domains and RTT Mutants. MBD = Methycl CpG Binding Domain; TRD = Transcriptional Repression Domain; CTD = C-Terminal Domain
88
EXPERIMENTAL PROCEDURES
Protein purification - Histone octamers were expressed and purified from bacteria, both with and
without tails as described (Luger et al., 1997). Recombinant human full length MeCP2 (isoform
e2, 486 amino acids) and the RTT mutants, MeCP2 (R133C) and MeCP2 (168X), were purified
as described (Georgel et al. 2003). Purified MeCP2 proteins were dialyzed against either HEGN
buffer (10 mM Hepes, 0.25 mM EDTA, 10% glycerol, and 2.5 mM NaCl) or H5EGN buffer (5
mM Hepes, 0.25 mM EDTA, 10% glycerol, and 2.5 mM NaCl). Glutathione S-transferase (GST)
fusion proteins were expressed and purified from Escherichia coli BL21 (DE3, pLys E) cells as
has previously been described (Georgel et al. 2001). Fusion proteins were constructed to include
one of the N-termini tails from each Drosophila melanogaster histone protein (H2A, H2B, H3, or
H4) or Green Fluorescent Protein (GFP).
Nucleosomal Array Reconstitution - The 208-12 DNA template was purified from the pPol-I-
208-12 plasmid (Georgel et al. 1993) as described (Hansen et al. 1989). Nucleosomal Arrays
were reconstituted using the 208-12 DNA template with histones containing the N-termini tails
(NAWT), without (NAtailless), and with all combinations (-H2A, -H2B, -H3, -H4) by using the salt
dialysis method (Hansen and Lohr 1993). Nucleosomal arrays were reconstituted at a ratio of 1.1
mol of histone octamer to mol of 208 bp of DNA. The final dialysis step was performed against
HEGN buffer for electrophoretic mobility shift assays and H5EGN for electron microscopy
experiments. Acetylated arrays were reconstituted using core histones treated with the histone
acetyltransferase GCN5 as previously described (Tse et al., 1998)
89
MeCP2-Nucleosomal Array Electrophoretic Mobility Shift Assays (EMSA) - Nucleoprotein
complexes were formed in the presence of increasing molar ratios of MeCP2 to 208 bp of DNA
repeat (rMeCP2 = 0.25-2.0) by incubation with arrays (50 ng in 15 µl) at room temperature for
15 min in HEGN buffer. Nucleoprotein complexes were analyzed by electrophoretic separation
for 1.5 hours at 10 V/cm on a 1 % agarose gel buffered with 1XTAE (40 mM Tris, 20 mM acetic
acid, and 1 mM EDTA, pH 8.3). Gels were stained with SYBR green (Invitrogen) at 1:10,000
dilution for 15 min.
GST pull-down assay and immunoblotting - GST and GST-fusion proteins (GST-H2A N-
terminal tail, GST-H2B N-terminal tail, GST-H3 N-terminal tail, GST-H4 N-terminal tail, and
GST-GFP) (2 µg) were incubated with MeCP2 (wt), MeCP2 (R133C), or MeCP2 (168X) (1 µg)
in the presence of BSA (10 µg) for 30 min at room temperature. Samples were mixed with a
50% slurry of glutathione beads (Amersham Pharmacia) incubated in TGD150 buffer (20 mM
Tris pH 8.0, 150 mM NaCl, 1mM DTT, 0.1% Triton X-100) with rotation at 4oC for 2 hours.
The beads were collected by centrifugation, and washed three times with TGD150. The
precipitated proteins were eluted in SDS-PAGE sample buffer, separated by SDS-PAGE, and
transferred to a nitrocellulose membrane. Membranes were probed with a MeCP2 antibody and
visualized using ECL according to the manufacturers protocol (Amersham Pharmacia).
Linker DNA accessibility assay - Nucleosomal arrays and MeCP2-nucleoprotein complexes
assembled as described for the EMSA analysis were incubated with 10 units of EcoRI/µg of
DNA for 90 min at 37oC. Digestion was stopped with the addition of EDTA to a final
concentration of 15 mM. The digested samples were treated with 10 µg of Protienase K for 60
min at 50oC, phenol chloroform extracted, and ethanol precipitated to determine the DNA
composition after digestion. The DNA was resuspended in 10 µl of 1X loading dye and analyzed
90
by electrophoretic separation on a 1% agarose gel at 8 V/cm for 2 h and stained with SYBR
green.
Electron microscopy- Transmission electron microscopy samples were prepared as previously
described (Nikitina et al., 2007). NAs and MeCP2 were mixed at the desired concentrations in
H5EGN buffer and incubated at room temperature for 30 minutes. The samples where then fixed
with 0.1% glutaraldehyde for 4 hours at 4oC before overnight dialysis into H5EGN, and applied
to glow-discharged carbon coated grids. The grids were positively stained with 1% aqueous
uranyl acetate followed by a thorough washing. These grids were subsequently examined in a
Tecnai 12 TEM operated at 100 KV in the tilted darkfield mode. The digital images were
captured by a TVIPS 2024x2024 CCD camera.
RESULTS
Histone N-terminal Tails Influence MeCP2-Chromatin Interaction- To understand the
importance of the N-terminal tails of histones on MeCP2-chromatin binding and/or
condensation, nucleosomal arrays were reconstituted using bacterially expressed histones either
with (NAwt) or without N-terminal tails (NAtailless). The 208-12 DNA template used for these
reconstitutions contains 12-tandem repeats of the 208 bp 5S rDNA nucleosome positioning
sequence from Lytechinus variegates that has been extensively used for chromatin structure
studies (for review see Hansen, 2002; Luger and Hansen, 2005). Binding of MeCP2 to
nucleosomal arrays was initially analyzed by Electrophoretic Mobility Shift Assays (EMSA)
(Figure 3.2, A). We have previously used this method to demonstrate MeCP2’s ability to bind
and condense nucleosomal arrays into higher-order chromatin structures (Georgel et al., 2003).
After incubation with wild-type recombinant MeCP2 at molar ratios rMeCP2 of 0.5-4.5, complexes
91
composed of arrays reconstituted with either wild-type histones or tailless histones were
electrophoresed on native 1% agarose (Figure 3.2, A). We found the recruitment of MeCP2 to be
significantly affected by the removal of all N-terminal tails of histones (Figure 3.2A: NAwt vs.
NAtailless EMSA and Figure 3.2, B: EM images). Both an increase in free array (black arrow) and
a decrease in NA/MeCP2 complex formation (white arrow) at ratios rMeCP2 0.5-4.5 were observed
with the tailless arrays (Figure 3.2, B lanes 3-6) compared to the wild-type array (Figure 3.2, A
lanes 3-6). Differences in levels of chromatin condensation between arrays with and without tails
were viewed by Electron Microscopy (EM) (Figure 3.2, C). Arrays comprised of full length
histones displayed both inter- and intra- chromatin fiber interactions. Condensation within
sections of arrays is clearly visible in some of the fields of view (yellow arrows). In addition,
there are numerous intra-fiber interactions seen with MeCP2 and wild type arrays (yellow stars).
The tailless arrays appear to lose some MeCP2-mediated fiber interactions. This appears to be
the result of both a reduction in the amount of condensed arrays and in the number of intra-fiber
interactions.
92
Figure 3.2, A and B: MeCP2 binding is influenced by histone N-termini. (a) Binding of MeCP2 to 208-12 NA. Lane 1 contains a 1 kb+ DNA ladder. Lane 2 corresponds to 208-12 NA alone. Lanes 3-8 are MeCP2 binding to 208-12 NA at increasing rMeCP2 (ranging from 0.25 to 4.0 MeCP2 molecules per 208 bp of DNA). The black arrow indicates free, unbound NA. The white arrow indicates shifted complexes. (b) Binding of MeCP2 to tailless 208-12 NA. Lane 1 contains 1 kb+ DNA ladder and lane 2 has tailless 208-12 NA alone. Lanes 3-8 correspond to tailless 208-12 NA with increasing rMeCP2 ranging from 0.5 to 4.5.
93
Figure 3.2, C: EM imaging of MeCP2-NA complexes with and without histone N-termini tails. (Panels A-C) 208-12 NA alone. (Panels D-K) 208-12 NA with MeCP2 at rMeCP2= 0.5. MeCP2-NA complex formation is observed to some degree in the majority of fields (yellow arrows). Intra- chromatin fiber interactions (yellow stars) can also be discerned in addition to inter-fiber compaction. (Panels L-O) Tailless 208-12 NA alone. (Panels P-W) Tailless 208-12 NA with MeCP2 at rMeCP2= 0.5. An increase in uncomplexed arrays is observed in the absence of histone N-termini. In addition, a decrease in intra-fiber interactions also appears to accompany the removal of N-termini tails from NA.
Individual Histone N-terminal Tails Effect on MeCP2 Binding- To evaluate the contribution of
individual histone tails, the same experimental approach was taken using NA reconstituted with
three wt histone and one tailless in all four combinations (NA-H2A, NA-H2B, NA-H3, NA-H4).
EMSA using NA-H2B as a template (Figure 3.3, B) shows no significant change in the MeCP2-
NA complexes formed as indicated by mobility shift patterns similar to that of normal NA
(Figure 3.2, A). This strongly suggests that H2B tails do not significantly contribute to MeCP2
recruitment to the NA. Deletion of tails from histones H2A, H3, and H4 impedes the formation
of MeCP2/NA complexes (shift delayed at similar rMECP2, compare lanes 2 to 6 in panels A, C,
94
and D) in a manner similar to what was observed with the tailless nucleosome arrays (Figure 3.2,
B, lanes 2-6). In addition, the increase in free NA (see black arrows Figure 3.3) confirmed a
change in MeCP2 binding affinity for NA-H2A, NA-H3, and NA-H4.
95
Figure 3.3: MeCP2 binding efficiency to composite NAs lacking one type of histone N-termini. (a) Binding of MeCP2 to NAH2A-. (b) Binding of MeCP2 to NAH2B-. (c) Binding of MeCP2 to NAH3-. (d) Binding of MeCP2 to NAH4-. Lanes 1 contain 1 kb+ DNA ladder. Lanes 2 correspond to composite 208-12 NA alone. Lanes 3-8, composite 208-12 NA with an increasing rMeCP2 ranging from 0.5 – 4.5. Black arrows indicate an increase in free NA in the H2A-, H3-, and H4- composite arrays. The white arrow denotes the NAH2B- shifting patterns that are similar to wild type NA.
96
Direct Interaction Between MeCP2 with Histone N-Terminal Tails- To directly test the in vitro
interactions between MeCP2 and core histone tails, we used bacterially expressed GST-histone
N-termini fusion proteins (Hecht et al., 1995; Georgel et al., 1997) as a ligand for MeCP2
binding assay (Figure 3.4). MeCP2 GST-fusion tail complexes were pulled down with
glutathione-Sepharose beads, eluted, and resolved with SDS-PAGE. The separated proteins were
subsequently transferred to a nitrocellulose membrane and visualized by immunoblotting using
an anti-MeCP2 antibody. When compared against 10% of input protein (Figure 3.4, A, lane 1),
MeCP2 was found to directly interact with GST-H2A and GST-H3 fusion proteins in vitro
(Figure 3.4, lanes 3 and 5), and to a lesser extent with the GST-H4 fusion (Figure 3.4, A, lane 6).
No interaction between MeCP2 and GST-H2B was detected in this assay (Figure 3.4, A lane 4),
matching our EMSA results. GST alone, or fused with green fluorescent protein (GFP) were
used as negative controls. No non-specific interaction with MeCP2 was detected (Figure 3.4, A,
lanes 2 and 8). These in vitro IP experiments confirm the MeCP2 electrophoretic mobility results
further demonstrating direct interactions between MeCP2 and the N-terminal tails of the core
histones H2A, H3, and H4.
97
Figure 3.4: MeCP2 interacts with the N-termini of H2A, H3, and H4 in a GST-fusion pull-down assay.
Lanes 1 and 7 indicate 10% (200 ng) of assay input MeCP2. Lane 2, MeCP2 and GST alone. Lanes 3, 4, 5, and 6 correspond to MeCP2 with GSTdH2A, GSTdH2B, GSTdH3, and GSTdH4 respectively. Lane 8 contains MeCP2 and GST-GFP.
Effect of Histone acetylation on MeCP2 binding and chromatin compaction- MeCP2 has recently
been described to localize not only in chromatin regions overlapping with repressed genes but
also to active loci in the genome (Yasui et al., 2007, Chahrour et al., 2008). These actively
transcribed regions are most often associated with hyper-acetylated H3 and H4 histone
(Jenuwein and Allis, 2001). Individual hyper-acetylated nucleosomes did not appear to be a
better substrate for MeCP2 binding (Ishibashi et al., 2008), but no information was reported on
the effect of acetylation on NA folding. To determine the influence of histone N-termini
acetylation on MeCP2 binding and chromatin condensation, acetylated nucleosomal arrays were
assayed by EMSA and imaged by EM. These NAs were reconstituted using hyper-acetylated
histones generated by the in vitro treatment of chicken erythrocyte histones with a histone
acetyltransferase (GcN5) (Figure 3.5). When comparing the affinity for normal and
hyperacetylated NA at rMeCP2 from 0.25 to 2.0, no noticeable difference in binding was observed
(Figure 3.5, B), in agreement with the published results on mono-nucleosomes (Ishibashi et al.,
2008). This confirms that acetylation does not significantly hinder MeCP2 recruitment to
98
chromatin. However, EM images of normal NA and NAacetylated indicated a reduction in MeCP2
ability to compact acetylated chromatin (Figure 3.5, B).
Figure 3.5, A: Effects of histone acetylation on MeCP2 binding. Binding of MeCP2 to acetylated 208-12 NA. Lanes 1 and 7 contain a 1 kb+ DNA ladder. Lanes 2 and 8 contain 208-12 NA without MeCP2. Lanes 3-6, unmodified 208-12 NA with increasing rMeCP2 from 0.25 to 2.0. Lanes 9-12, acetylated 208-12 NA with increasing rMeCP2 from 0.25 to 2.0.
99
Figure 3.5, B: Effects of histone acetylation on MeCP2 compaction. EM imaging (a) acetylated 208-12 NA alone (b) MeCP2 and acetylated 208-12 NA complexes at rMeCP2 = 1.
RTT-causing Mutant MeCP2 R133C Interactions with N-Terminal Histone Tails- The mutant
MeCP2 R133C contains a point mutation at residue 133 converting an arginine to a cysteine
residue which has been linked to patients suffering from RTT (Amir et al., 1999). This point
mutation in the MBD has been shown to dramatically hinder the preference for MeCP2 to bind to
methylated DNA (Ballestar et al., 2000). Structurally this mutation is expected to prevents proper
interaction between the MBD and its DNA target (Ohki et al., 2001; Ballestar et al., 2000; Lee et
al., 2001; Van den Veyver et al., 2001). We have previously demonstrated this point mutation to
have no significant effect on binding to normal NA and to have a limited impact on the in vitro
formation of oligomeric superstructures (Georgel et al., 2003). To complement this study, we
decided to investigate the contribution of the individual histone tails to MeCP2 R133C
recruitment. Similar to what was done with the wt MeCP2, we performed EMSA with the
various combinations of composite NAs (Figure 3.6, A). Expectedly, the results indicate a
structural role for the tails in the binding mechanism of MeCP2 R133C (NAwt vs. NAtailless). But,
100
in contrast to the differential binding observed with the wt MeCP2 (no significant H2B tail
contact), the R133C mutant appears to be able to recognize all four histone N-termini with nearly
equivalent affinity. When comparing the results obtained using normal NA (Figure 3.6, A, lane
3), the removal of any individual tail (including that of H2B) resulted in a decrease in MeCP2
R133C binding efficacy to composite NA, as indicated by the remaining uncomplexed free NA
at rMeCP2 = 0.25 (Figure 3.6, C-F, lane 3). The lack of histone tail discrimination by the MeCP2
R133C results were confirmed by co-IP using the GST-tail constructs previously described
(Figure 3.6, B). The relative signal intensity for GSTdH2A, GSTdH3 and GSTdH4 were
comparable to that observed with wtMeCP2 (Figure 3.4). The difference in function resides in
the acquired ability for the MeCP2 R133C to recognize and bind the N-terminal region of H2B.
This result suggests that the R133C mutation may induce a conformational change at the
secondary or tertiary protein folding level.
101
Figure 3.6, A: RTT-linked MeCP2 R133C differential chromatin interaction. (Panel A) MeCP2 R133C with unmodified 208-12 NA. (Panel B) MeCP2 R133C with tailless 208-12 NA. (Panel C) MeCP2 R133C with 208-12 NAH2A-. (Panel D) MeCP2 R133C with 208-12 NAH2B-. (Panel E) MeCP2 R133C with 208-12 NAH3-. (Panel F) MeCP2 R133C with 208-12 NAH4-. Lanes 1 contain 1 kb+ DNA ladder. Lanes 2 correspond to 208-12 NA alone. Lanes 3-7, 208-12 NAs with increasing rMeCP2-R133C from 0.25 to 2.0.
102
Figure 3.6, B: GST-fusion pull-down with MeCP2 R133C. Lanes 1 and 7 indicate 10% (200 ng) of assay input MeCP2 R133C. Lane 2 is MeCP2 R133C and GST alone. Lanes 3, 4, 5, and 6 correspond to MeCP2 R133C with GSTdH2A, GSTdH2B, GSTdH3, and GSTdH4 respectively. Lane 8 contains MeCP2 R133C and GST-GFP.
The Chromatin-condensing ability of MeCP2 Is Linked to Histone Tail Binding- The RTT-linked
R168X MeCP2 possesses a mutation that alters the arginine 168 codon to a stop codon, leading
to a truncated protein that only contains the N-terminus and MBD domain (Lee et al., 2001; Van
den Veyver et al., 2001). Previously we have reported this RTT mutant to have a significantly
decreased binding affinity for NA when assayed by native agarose gel electrophoresis (Georgel
et al., 2003). This mutation does not prevent MeCP2-NA interaction, but leads to a loss of
function in NA condensation and an inability to promote the formation of larger oligomeric
suprastructures. This study suggested MeCP2s ability to condense chromatin into higher-order
structures occurred through its transcriptional repression domain or other C-terminal domains.
The binding pattern observed with MeCP2 R168X and NAwt (Georgel et al., 2003) is nearly
identical to that of wtMeCP2-NAtailless binding (Figure 3.2, B), suggesting this region of
MeCP2 is necessary for histone tail contact and is also critical for the induced chromatin
compaction. To confirm that the EMSA patterns are the result of an inability for MeCP2 R168X
to recognize the histone tails, we performed co-IP experiments under conditions similar to that
103
described for the wt and R133C proteins. These results showed that deletion of the residues
outside of the N-terminus and MBD results in the complete loss of histone tail interactions
(Figure 3.7, lanes 3, 4, 5, and 6). This confirms that histone tail interactions require the presence
of domains other than the MBD, such as the TRD, CTDα, or CTDβ.
Figure 3.7: The RTT-linked truncated MeCP2 R168X mutant loses histone N-termini interaction. GST-fusion pull-down assay with RTT-linked MeCP2 R168X. Lanes 1 and 7 indicate 10% (200 ng) of assay input MeCP2 R168X. Lane 2 is MeCP2 R168X and GST alone. Lanes 3, 4, 5, and 6 correspond to MeCP2 R168X and GSTdH2A, GSTdH2B, GSTdH3, and GSTdH4 respectively. Lane 8 contains MeCP2 R168X and GST-GFP.
DISCUSSION
Characterization of Human MeCP2-Chromatin Binding Determinants- The experiments
described in this study were designed to elucidate human MeCP2’s chromatin binding
determinants and evaluate their influence on chromatin condensation. MeCP2 has long been
described as a global repressor that is localized to methylated regions of the genome (Nan et al.,
1997). Scientific interest in this protein increased dramatically after MeCP2 mutations were
linked to a severe neurodevelopmental disorder, the RTT (Amir et al., 1999). In addition to
MeCP2 preferentially binding methylated DNA over unmethylated DNA, a significant role in
chromatin compaction ability has recently been demonstrated (Georgel et al., 2003). This in
vitro binding study used unmethylated DNA incorporated into chromatin as a template,
104
separating MeCP2’s methylated DNA binding preference from chromatin condensing function.
In order to further characterize MeCP2-chromatin interactions, we assembled arrays of
nucleosomes that lacked all or one of the four N-terminal tails of the core histones H2A, H2B,
H3, and H4. Our in vitro binding assays surprisingly displayed a significant differential MeCP2
binding between normal chromatin and its tailless counterpart, as demonstrated by the increase in
free NAtailess versus NAwt (Figure 3.2, lanes 4 and 5). These results indicate a clear influence of
histone N-termini in MeCP2-chromatin interactions. This decreased affinity is also accompanied
by a loss of efficiency in the formation of MeCP2-mediated higher order chromatin complexes.
The cleavage of linker DNA by restriction enzymes was not sufficient to entirely destabilize the
MeCP2-NA complexes (Georgel et al., 2003). This study in combination with previous work that
showed MeCP2 exists as a monomer (Klose and Bird, 2004) strongly suggests a stabilizing
effect through the formation of bivalent DNA-MeCP2-DNA “bridges” between nucleosomes.
The reduction in MeCP2-NA complex formation in the absence of histone tails suggests these
“bridges” may involve histone-N-termini in MeCP2 recruitment and/or complex stabilization.
The EM images (Figure 3.2) reflect reduced interactions between MeCP2 and the NA in the
absence of histone N-termini. The effect was observed, both in cis (intra-NA) and trans (between
adjacent NAs), indicating that recruitment and formation of supramolecular complexes were
affected. The absence of some specific long-range chromatin interactions (kilobases apart) in
MeCP2-null mice suggests an in vivo role for MeCP2-mediated bridges in higher order
chromatin loop structures (Horike et al., 2005).
To date, there is no evidence that directly links MeCP2 function to the histone tails in the
context of chromatin. Cross-linking experiments have previously suggested that MeCP2 binding
occurs in the vicinity of histone H3 (Nikitina et al., 2007). Subsequent structural experiments
105
performed on mono-nucleosomes clearly showed the histone H3 tail not to be critical for MeCP2
binding (Ishibashi et al., 2008). Here, composite nucleosomal arrays assembled using three wild
type histones combined with one tailless histone were used to more clearly delineate the
individual contribution of each of the N-termini region on MeCP2-NA condensation. EMSA
results associated the reduced ability of MeCP2 to mediate complex formation with the absence
of individual N-terminal tails of histones H2A, H3, and H4 (Figure 3.3). In contrast, the loss of
H2B N-terminus did not affect MeCP2 binding. The difference in H3 tail interactions reported
between single nucleosomes and NA may result from an intra-nucleosomal array effect where
the N-terminal region contributes additively and independently to NA oligomerization (Dorigo et
al, 2004; Zheng et al, 2005; Gordon et al, 2005; Arya and Schlick, 2006). This may indicate two
or more adjacent N-terminal tails stabilizes MeCP2 chromatin contacts. The contribution of
MeCP2-histone tail interaction may manifest itself in regulating compaction of larger arrays of
nucleosomes and therefore no evident effect would be detected when using mono-nucleosome
for binding studies (Ishibashi et al., 2008). Computer modeling analyzing the role of individual
tails on chromatin compaction as a function of Mg2+ concentration predicts the tails of H3 and
H4, but not H2A or H2B, play a significant role in NA condensation (Arya and Schlick, 2009).
In order to differentiate between core histone N-termini tails involvement in promoting
chromatin folding vs. directing histone N-termini interaction by MeCP2, GST-histone tail fusion
Co-IPs were performed. These GST-fusion proteins have previously been used to demonstrate in
vitro binding capabilities of Sir3, another CAP (Georgel et al., 2001). Here, these constructs
were used to demonstrate the direct binding of MeCP2 to the histone N-terminal tails of H2A,
H3, and H4 by this Co-IP assay (Figure 3.4). The results of these pulldown experiments were
consistent with the EMSA conclusions which indicated a lack of interactions between the H2B
106
tail and MeCP2. The absence of MeCP2-H2B tail interactions questions the role H2B N-termini
may play during MeCP2 gene regulation. One possibility is by leaving the H2B tail sterically
unhindered, it would be exposed to PTM signaling and/or other CAP interactions. Identification
of MeCP2-histone tail interaction strongly suggests the histone code may play a role in MeCP2
function and/or localization. These results when taken together indicate MeCP2 directly interacts
with and is aided by the N-termini of histones to promote chromatin folding.
Acetylation Influences MeCP2-mediated Chromatin Condensation- Based on existing literature
describing MeCP2 as a general co-repressor, interactions with acetylated histone tails are
expected to be limited. MeCP2 has classically been linked to histone hypoacetylation as
evidenced by its co-purification with Sin3-histone deacetylases complexes (Jones et al., 1998;
Nan et al., 1998). However, recent work has also linked MeCP2 to actively transcribed loci,
leaving the potential for MeCP2 to act as a co-activator. A possible role for MeCP2 as a co-
activator came from the recent findings of MeCP2 association with promoters of active genes
(Yasui et al., 2007, Chahrour et al., 2008). In addition to interactions with Sin3-HDAC
complexes, in vitro preference of MeCP2 for methylated DNA further indicates a role as a co-
repressor. However, recent studies have indicated that DNA methylation may not be required for
MeCP2 recruitment to chromatin and induced compaction (Georgel et al., 2003). The proposed
mechanism of preferential binding to methylated DNA and chromatin folding contrasts with the
recent findings of MeCP2-associated transcriptionally active loci. Based on such evidence for a
dual role as co-repressor and co-activator, MeCP2 is now considered as a transcriptional
regulator whose recruitment to regulatory DNA elements and subsequent function is likely
mediated by histone PTMs. Acetylation of mono-nucleosomes has been previously been
demonstrated to not interfere with the in vitro binding of MeCP2 (Ishibashi et al., 2008). Our
107
own in vitro acetylated NA experiments confirmed these results, showing little difference in
binding efficiency when compared to control NA (Figure 3.5, A). Interestingly, we do find
histone acetylation to impede the condensation of chromatin as evidenced by EM images
showing only partial MeCP2-induced condensation of the acetylated NA (Figure 3.5, B). The
effect may simply be due to an inability for MeCP2 to overcome the electrostatic repulsion
generated by the acetyl groups present on the histone N-termini. The apparent inability for
MeCP2 to differentiate between normal and hyper-acetylated tails is compatible with both co-
repressor and co-activator roles. The difference in biological function may lie in MeCP2
interactions with other co-repressors or co-activators. The interaction with the Sin3-HDAC
complexes would definitely favor MeCP2’s ability to compact chromatin and therefore further
repress loci that are poised for repression. The switch from co-repressor to co-activator or vice
versa may be the result of differential post-translational modifications of MeCP2. (Zhou et al.,
2006; Tao et al., 2009). Another potential consequence of MeCP2 binding to transcriptionally
active loci may be mediated by chromatin folding. Despite a lower ability to compact chromatin
when histone tails are hyper-acetylated, MeCP2 can still contribute to folding. This partial
folding may induce secondary structures that are conducive to transcription (Georgel et al.,
2003). The reduced chromatin compaction may also expose unique unidentified MeCP2 binding
determinants critical for the recruitment of other CAPs interacting over long distances.
Additionally, in vivo localization of MeCP2 has been demonstrated to coincide with an
increase in H3K9 methylation on the repressor domain in the upstream regulatory region of the
H19 gene (a MeCP2-regulated gene (Drewell, et al., 2002)) (Fuks et al., 2003). This correlation
was suggested by the authors to be caused by the tethering of an unidentified histone
methyltransferase to the N-terminal region of MeCP2. The presence of methylated H3K9 by
108
itself may change MeCP2’s affinity to chromatin, as well as its folding properties. The binding
of MeCP2 to H3 tail may also enhance the methyltrasnferase activity of this unknown enzyme by
properly orientating the histone tail. The chromatin compaction ability of MeCP2 appears to be
influenced by the presence of histone tails and their PTM, but a further characterization of the N-
termini PTMs role in MeCP2 recruitment and/or function will have to be performed as the
availability of specifically modified histone increases.
Role of RTT Mutations on Chromatin Interaction - To determine the influence of RTT mutations
on recruitment to chromatin templates, we assessed the differential chromatin binding properties
of two RTT-linked MeCP2 mutants (R133C and R168X). The R133C MeCP2 mutation alters
one of the five residues that generate the hydrophobic methyl-binding pocket of the MBD (Free
et al., 2001; Wakefield et al., 1999). This mutated form of MeCP2 displays a 100-fold less
preference for methylated DNA compared to wild type MeCP2 (Ghosh et al., 2008). Previously
published binding studies using MeCP2 R133C and 208-12 NA showed this mutation to have
little effect on chromatin interactions (Georgel et al., 2003). Surprisingly, our EMSA binding
studies identified a differential binding pattern associated with the R133C MeCP2 mutant and
the histone H2B tailless NA-H2B (Figure 3.6, D). This gain of function was confirmed by our
GST-fusion immunoprecipitation experiments (Figure 3.6, G, lane 4). The additional ability for
MeCP2 R133C to recognize and bind H2B N-terminus may play a role in RTT development of
specific cell types. This may be explained through steric hindrance preventing the recruitment of
other specific factors that would otherwise interact with the H2B tail, resulting in deficient CAP
binding or signaling that would impact proper regulation of specific genes leading to various
developmental defects.
109
The additional H2B tail binding could change the MeCP2-mediated chromatin structure
normally adopted by the wild type protein. Alterations in chromatin structure could influence
either of MeCP2s co-repressor or co-activator activities. These point mutations gain of H2B
binding also could affect MeCP2 recruitment to a small number of critical developmental genes,
but not grossly change the global localization of this RTT mutant. It is also possible that the
wtMeCP2 can bind to H2B N-termini, but only in the presence of a specific PTM, with this point
mutation removing PTM selectivity. In order to determine the region of MeCP2 responsible for
histone tail interaction, the RTT R168X truncated mutant was analyzed with our GST-fusion
immunoprecipitation assay. The MeCP2 R168X truncation has previously been used to
demonstrate the area responsible for MeCP2-mediated higher order chromatin structure
formation lies outside of the N-terminus and MBD domain (Georgel et al., 2003). Our GST-
histone tail fusion pull-downs with MeCP2 R168X co-localize the loss of histone tail interaction
with the loss of higher order chromatin compaction to the TRD or other C-terminal residues
(Figure 3.7). We suspect that these two MeCP2 functions may be interlinked, with MeCP2
condensing chromatin into higher order structures through a DNA/histone tail-MeCP2-
DNA/histone tail “bridging” mechanism. Additional studies, possibly through additional
truncated versions of MeCP2, will have to be performed to separate these two MeCP2 functions
in an attempt to further elucidate their interconnection.
Curiously, the R133C mutation of MeCP2 resides in the MDB region, well within the
first 168 aa that we demonstrate to not interact with histone N-termini, but was found to
influence histone H2B tail binding. MeCP2 has been described to have an abnormal tertiary
structure compared to other identified nucleoproteins (reviewed in Hite et al., 2009). It was first
characterized to have an anomalous molecular mass when analyzed by gel filtration (~70-80
110
kDa) yielding a different mass than calculated based on residue sequence (52.4 to 53 kDa)
(Meehan et al., 1992; Klose and Bird, 2004). The structure of MeCP2 has subsequently reported
to be intrinsically disordered (Adams et al., 2007). This study demonstrated almost 60% of
MeCP2s tertiary organization to be unstructured. The large quantity of unstructured regions has
been hypothesized to be responsible for preventing the crystallization of full-length MeCP2. The
arginine to cysteine point mutation of the R133C mutant adds a sulfide group that could mediate
a possible additional disulfide bond in the mutant form but not in the wt MeCP2 protein. This
additional disulfide bond could alter the secondary and tertiary structure of MeCP2 and/or
change regions involved in histone N-termini interactions. Supporting this possibility, changes in
secondary structure have been reported for the R133C, as well as T158M, R106W, and F155S
correlating with differences in methylated DNA binding (Ballestar et al., 2000). These
conformational changes may additionally have an effect on either chromatin binding or MeCP2-
associated chromatin compaction ability. Furthermore, the truncation of MeCP2 R168X may
affect the overall folding in such a way that could explain the loss of histone tail interactions
observed in our GST-fusion immunoprecipitation experiments.
These studies on identification of differential chromatin interaction amongst RTT-linked
MeCP2 mutants may lay the groundwork for a more in depth identification of the molecular
basis of RTT. The next step in MeCP2 characterization will require a comprehensive analysis of
histone N-termini tail PTMs and their effect on chromatin binding and condensation. The
characterization of PTMs involvement in wtMeCP2s localization and function needs to be
determined because some differences in RTT function may involve the improper reading of
specific PTM or sequence of PTMs. The differential binding of both RTT MeCP2 mutants used
in this study strongly indicates that other RTT-linked mutations may also be influencing MeCP2-
111
chromatin interactions in a histone tail-mediated manner. The RTT linked differences in
chromatin binding could begin to explain the dysfunctional molecular mechanism of RTT, and
we believe will become one of the main foci of future MeCP2 studies.
112
CHAPTER 4: DISCUSSION, CONCLUSION, AND FUTURE STUDIES
DISCUSSION
Additional Chromatin Interactions of Eukaryotic Gene Silencing CAPs
In recent years, documentation regarding the effect that chromatin structure may have on
gene regulation has significantly increased. However, the specific role and mechanism of the
individual chromatin components, their regulatory functions, and interacting partners are still
only partially understood. This dissertation addresses the function and interacting partners of two
well studied CAPs: Sir3 and MeCP2. The Sir3 protein in Saccharomyces cerevisiae has
classically been described for its gene silencing activity through recruitment by the Rap1 protein
and subsequent interactions with the de-acetylated N-terminal unstructured tails of histones H3
and H4. In contrast, the mammalian MeCP2 protein has long been characterized to repress
transcription through its selective binding of methyl-CpG DNA. The studies performed here
have identified and characterized additional chromatin interactions involving Sir3 and MeCP2,
and demonstrated that these interactions may hold significant biological implications. In
addition, this work challenges the previous mode of action of these two CAPs, leading us to
revisit their role in the context of chromatin interactions. While Sir3 and MeCP2 do not contain a
classical DNA-biding domain or histone interacting domain respectively, they are proteins with a
high intrinsic disorder, a property often associated with proteins involved in multiple protein-
protein and protein-DNA contacts (reviewed in Uversky et al., 2005). This chapter will highlight
the biological significance of my research findings, postulate further on the molecular
mechanism of these CAPs, and describe future experiments that could provide additional
evidence for novel and important functions of Sir3 and MeCP2.
113
Role of Nucleic Acid Binding in Sir3-Dependent Interactions with Chromatin Fibers
The silencing protein Sir3 has recently been characterized to bind to the DNA fiber
within chromatin, in addition to its interaction with histone H3 and H4 N-termini (Georgel et al.,
2001). This initial work strongly suggested that Sir3 would bind in a similar manner to native
template chromatin (208-12 NA), tailless arrays, and naked DNA. Our experiments confirmed
these initial observations (Figure 2.1), and further characterized the role of nucleic acids binding
by Sir3 in order to further elucidate the role of Sir3-chromatin interactions.
In order to determine the DNA features within chromatin which may preferentially
interact with Sir3, we examined the effect of DNA length (Figure 2.2) and structure (Figure 2.3)
on Sir3 binding. We found that the binding ability of Sir3 was influenced more by changes in
DNA structure than by size variation. The more “straight” fragment (HSP promoter) was a better
substrate for Sir3 binding, followed by the “bent” (pPol I promoter), then “curved” (208-1 DNA)
fragment (Figure 2.3). The structures of these three DNA fragments were modeled with software
based on the Trifonov algorithm (Figure 2.3, A) designed to predict substrate curvature based on
sequence. Structural changes were confirmed and characterized by Quantitative Agarose Gel
Electrophoresis (Table 2.1), allowing us to calculate the radii of the three substrates to confirm
the differential curvature. This quantifiable preference for a “straight” DNA structure initially
indicated that linker DNA within chromatin is a more likely Sir3 binding partner, as opposed to
“curved” nucleosomal DNA. This is due to the structure of linker DNA being expected to display
a more straight conformation as opposed to nucleosomal DNA which possesses an induced
curvature mediated by core histones interactions.
114
In order to test whether the short stretch of DNA at the entry/exit point of the nucleosome
was a target for Sir3 binding, a “4-way junction” DNA template comprised of 4 oligonucleotides
annealed in a specific order was used as a substrate for our standard binding assay. This
construct has previously been used to mimic DNA cross-over targeted by other architectural
CAPs such as HMGA proteins, HMGB proteins, and linker histones (Ferrari et al., 1992;
Panyutin and Hsieh, 1993; Hill and Reeves 1997; Varga-Weisz et al., 1994). We found that the
4-way junction DNA was the weakest DNA substrate tested in all of our Sir3 binding assays,
indicating Sir3-DNA interactions are weak at the nucleosomal entry-exit point. This result is in
contrast with previous studies using a DNA accessibility assay on Sir3-NA complexes which
demonstrated no significant blockage of linker DNA digestion (Georgel et al., 2001). Another
possible alternative for Sir3-DNA contacts may occur at the dyad DNA section of nucleosomes.
This area, between nucleosome base position 60 and 90, displays a straighter DNA
conformation, as opposed to the remaining nucleosomal DNA. As such, it may provide a
preferred DNA structure for Sir3 binding. Our length dependent binding study demonstrated that
Sir3 could complex with DNA fragments as short as 12-bp in length, supporting the possibility
that Sir3-DNA contacts can occur within nucleosomal DNA or short stretches of linker DNA
(Figure 2.2).
Throughout our nucleic acid characterization, Sir3 was found to cooperatively bind to
nearly all fragments assayed by EMSA (Figure 2.3 and 2.4). This cooperative binding is
demonstrated by an abrupt loss of free DNA when Sir3 reaches the critical DNA-binding
molarity to form complexes. The mass of free DNA was measured at increasing Sir3 molar ratios
and plotted (Figure 2.3, C and Figure 2.4). The quantification of uncomplexed DNA was
performed as a function of fluorescence intensity by matching quantified signal against a serial
115
dilution of free DNA. These results indicated a nearest-neighbor cooperative mode of binding
with the plotted data demonstrating a fitted curve with a distinct S-shape as opposed to a linear
plot associated with sequential independent binding. Imaging by EM supported the cooperative
binding model for Sir3-DNA interaction. Large Sir3-DNA complexes were observed adjacent to
free DNA in the same field of view (Figure 2.5, panels D, E, and F). In addition to Sir3
displaying a cooperative binding activity to naked DNA templates, a concentration dependent
effect on binding was also observed (Figure 2.6). An increase in binding by doubling the number
of molecules while keeping the same molar ratio between protein and DNA fragments was found
to occur with native chromatin templates (Figure 2.6, A), the tailless array (Figure 2.6, B), as
well as naked DNA (Figure 2.6, C). This observation matches what has been described in vivo,
where the level of Sir3 protein in yeast was found to directly impact the spreading of silencing
from the telomeric regions in a dosage dependent manner (Renauld et al., 1993). These
combined observations support a model where the spreading of genetic silencing in yeast by Sir3
would occur through cooperative binding to DNA as a nucleating event. This is then followed by
the subsequent cooperative recruitment of other Sir3 molecules, resulting in spreading of
heterochromatin and an increase in the position effect variegation of nearby genes.
Unexpectedly, we discovered that not only can Sir3 bind to double-stranded DNA, it can
also complex with single-stranded DNA in vitro (Figure 2.7). The DNA at telomeres exists in a
single-stranded form where in yeast the single-stranded telomeric DNA-binding protein, Rlf6, is
known to bind and effect the localization of Rap1 (Konkel et al., 1995). Sir3 has been known to
localize to silent areas in the yeast genome, including telomeric regions (Aparicio et al., 1991),
and interact with Rap1 and Sir4 in vivo to maintain silencing at these loci (Rine and Herskowitz,
1987; Kyrion et al., 1992; Gotta et al., 1996). Our data suggests that the Sir3 protein through this
116
single-stranded DNA binding capability may be involved in a mechanism for stabilizing the
Rlf6-Rap1 complex at single-stranded telomeric DNA regions.
Role of Histone N-termini in MeCP2-Chromatin Interactions
MeCP2 has classically been defined as a methyl-DNA binding silencing protein, but has
recently been shown to display additional activities linking it to both activation of gene
expression and RNA splicing (Young et al., 2005; Yasui et al., 2007; Chahrour et al., 2008). In
vitro MeCP2-chromatin condensation studies have also demonstrated MeCP2-mediated
chromatin compaction to be independent of DNA methylation specific binding (Georgel et al.,
2003). In light of these findings, we set out to determine if the N-termini tails of core histones
influence MeCP2-chromatin interactions. Bacterially-expressed tailless histones were purified
and used for chromatin reconstitution into 208-12 NA. These NA were then incubated with
MeCP2 at differing molar ratios and complex formation was assayed by EMSA (Figure 3.1, B)
and EM (Figure 3.1, C). The EMSA studies revealed that the removal of histone N-termini had a
significant effect on MeCP2 recruitment. Higher molar ratios of MeCP2 were required to achieve
tailless NA complex formation similar to their normal NA counterparts (Figure 3.1, A vs. B).
EM imaging revealed a corresponding decrease in MeCP2-mediated chromatin interactions and a
decrease in the level of condensation associated with the removal of histone N-terminal tails
(Figure 3.1, C). This is the first evidence which suggests that MeCP2 interacts with histone N-
termini, and it opens the possibility of another level of MeCP2 recruitment through specific
histone modifications. There is some supporting in vivo data linking the localization of MeCP2
with H3K9 methylation at observed loci in mouse fibroblast cells (Fuks et al., 2003), perhaps
indicating a preference for this modification in the process of MeCP2 binding.
117
To determine if individual tails have an influence on MeCP2-chromatin interactions,
composite arrays comprised of three wild type and one tailless histone were reconstituted and
incubated with increasing amounts of wt MeCP2 (Figure 3.2). While removal of individual tails
did not affect binding to the same degree observed with the tailless arrays, MeCP2-complex
formation was reduced in arrays lacking individual H3, H4, and H2A histone N-termini (Figure
3.2, A, C, and D). MeCP2 complex formation was essentially unaffected by the lack of H2B
tails, as shown by the EMSA profiles (Figure 3.2, B). Our EMSA results indicate a histone tail
effect on MeCP2-mediated chromatin condensation. From these results, we cannot unequivocally
conclude whether the MeCP2-mediated chromatin compaction follows the same rules as the
standard histone H3 and H4 tail-mediated compaction described in other chromatin structural
studies (Zheng et al., 2005; Kan et al., 2007; Kan et al., 2009). GST-fusion histone N-termini
were used to assess MeCP2-histone N-termini interactions providing additional information on
this CAPs chromatin condensation ability. These GST constructs have previously been used to
investigate other CAP-histone N-terminal interactions (Georgel et al., 2001). Our GST-pulldown
experiments revealed a direct interaction between MeCP2 and histone H2A, H3, and H4 N-
termini (Figure 3.3). These results confirm that direct MeCP2-histone tail interactions occur in
vitro, and may be influencing MeCP2-mediated chromatin condensation.
Direct contact between histone tails and MeCP2 within chromatin signifies a possible
role for histone post-translational modifications in MeCP2 binding, localization, and/or function.
Since MeCP2’s recently described association with promoters of active genes (Yasui et al., 2007;
Chahrour et al., 2008), where CpG methylation may be at a minimum, a novel mechanism of
MeCP2 recruitment should be envisioned. Since histone tails appear to play an important role,
their post-translational modifications may further modulate MeCP2 recruitment. To test for the
118
effect of acetylation, one of the most common PTMs, we examined the influence of
hyperacetylated histone tails on MeCP2-mediated chromatin condensation (Figure 3.4). While
we found histone acetylation to not have an effect on MeCP2 complex formation compared to
un-modified NA (Figure 3.4, A), EM imaging revealed a reduction in induced chromatin
condensation (Figure 3.4, B). This result matches the predicted function of MeCP2 as it is known
to interact with Sin3/HDAC1 at specific target loci (Nan et al., 1998; Suzuki et al., 2003). These
findings suggest MeCP2 recruitment to chromatin is not completely hindered by acetylated N-
termini tails, which support the previous finding that MeCP2 can function not only as a
transcriptional repressor, but is also capable of associating with active genes marked by histone
acetylation (Yasui et al., 2007; Chahrour et al., 2008). The role of MeCP2 at the promoters of
these active genes is currently unknown, but may be linked to the modulation of local histone
arrangement into distinct chromatin structures (Georgel et al., 2003). The role of MeCP2 on
active genes may also be the result of long-distance genetic interactions. As MeCP2 induces
chromatin folding, it may bring distant regulatory elements into close proximity through a
potential looping mechanism, as has been postulated for its silencing function (Horike et al.,
2005).
The role of MeCP2 and how mutations impact its function in the underlying molecular
mechanism of RTT development remains poorly characterized. Here, we examined two common
RTT mutations, one with a single-point mutation in the MBD, MeCP2 R133C, and one with a
point mutation resulting in an early stop codon, generating a truncated mutant, MeCP2 R168X
(Amir et al., 1999). The MeCP2 R133C mutation has previously been described to have lost its
selectivity for methyl DNA (Usufzai and Wolffe 2000), yet retains its ability to condense
chromatin (Georgel et al., 2003). Our GST-fusion study demonstrates MeCP2 R133C point
119
mutant to be a gain of function mutation. As the wt MeCP2 can only interact with histones H2A,
H3, and H4 (Figure 3.5, B), the R133C mutant can recognize all four histone tails. This
additional tail interaction may signify an unnatural point of contact interfering with proper
MeCP2-mediated chromatin condensation. It also may indicate that wt MeCP2 can only
recognize H2B N-termini when it is adequately post-translationally modified. The point mutation
may simply abolish this selective binding. Additionally, EMSA experiments determined the
R133C point mutation had a reduced binding with composite arrays comprised of H2B tailless
histones (Figure 3.5, A) as opposed to the wt MeCP2 being affected by the loss of only H2A, H3,
and H4 tails. However, this difference in mobility with removal of the individual tails was
observed at a lesser degree than that in the wt MeCP2 composite arrays EMSA (Figure 3.3). To
begin to delineate the region of the MeCP2 protein responsible for the histone tail interactions,
the truncated mutant, MeCP2 R168X, was assayed using our GST-fusion histone N-termini pull-
down assay (Figure 3.6). This experiment revealed that the mutant lacking the TRD and CTDs
did not bind to any of the four histone tails. Since the MeCP2 R168X mutant was previously
demonstrated to retain partial binding to 208-12 template chromatin, it appears that MeCP2 can
contact nucleosomes through determinants other than the N-terminal tails. However, the R168X
mutant was not capable of generating higher order structures of the type observed with wt
MeCP2 (Georgel et al., 2003). In conclusion, our pull-down experiments suggest that some of
the previously characterized effects of MeCP2 on generating chromatin higher order structures
could be mediated by histone N-termini contacts. The changes in MeCP2’s ability to condense
chromatin in a histone acetylation-dependent manner indicate a role for post-translational
modifications. These potential regulatory events will require further research to clearly identify
the critical histone tail modifications involved in MeCP2-chromatin interactions. Because these
120
two RTT mutants differentially bind to core histone tails, it appears that the histone code, or
maybe more accurately the lack of proper recognition by mutant MeCP2, may play a crucial role
in the molecular mechanism of RTT development.
Possible Role of Intrinsic Disorder in Sir3/MeCP2 Chromatin Interactions
In addition to structured domains defining protein’s function and partners, a large number
of them contain a significant amount of unordered segments, as demonstrated by comparison of
the eukaryotic proteome (Oldfield et al., 2005; Liu et al., 2006). These “unorganized” segments
are known to be unable to adopt standard three-dimensional secondary structures and are referred
to as being intrinsically disordered (reviewed in Uversky et al., 2005). Protein intrinsic disorder
(ID) has been explored for many years, and has been referred to by many terms such as “proteins
being partially unfolded” (Linderstrøm-Lang and Schellman, 1959), flexible (Pullen et al., 1975),
or mobile (Cary et al., 1978). The higher amounts of intrinsic disorder found in eukaryotic
proteins have been theorized to correspond to the increased complexity required for additional
protein-protein interactions important in gene regulation and cell signaling. These multimeric
complexes may not be necessary in less complex organisms. In support of this theory, a
comprehensive analysis of eukaryotic transcription factors has identified that a large majority of
them contain extended intrinsically disordered regions (Liu et al., 2006). These long stretches of
ID regions have been proposed to participate in the formation of a flexible scaffold required for
multiple biological interactions. To support this theory, multiple studies have linked these
disordered regions of proteins with a large number of functions, including DNA interactions
(Spolar and Record 1994; Weiss et al., 1990; Paull et al., 2001), RNA-binding (Allain et al.,
1996; Markus et al., 1997; Nanduri et al., 1998), and protein-protein associations (Muro-Pastor et
al., 2003; Scully et al., 1997; Zhong et al., 1999; Zhang et al., 1998). Although these intrinsically
121
disordered regions lack a detectible structure, some areas have been described to undergo
conformational changes that lead to the adoption of ordered structures upon binding to their
physiological partner (Li et al., 2000; Bothner et al., 2001; Bourhis et al., 2004). The lack of a
characterized, ordered domain responsible for Sir3’s DNA binding ability could point to the
large disordered regions within this CAP as responsible for this interaction. Similarly, MeCP2’s
ability to recognize histone N-termini may be related to its high levels of intrinsic disorder.
The Sir3 protein has recently been described to bind to DNA in vitro, despite the lack of
an identifiable DNA binding domain. Throughout previous studies and during our Sir3-nucleic
acid characterization, no sequence dependent binding has been observed for this CAP. Primary
amino acid sequence analysis of Sir3 reveals some basic charge patches (Green highlight, Figure
4.1) that may indicate potential DNA contacts. Analyzing the primary sequence by predictor of
natural disorder regions (PONDR) VL-XT, one of the algorithms used to identify IDs, also
reveals a high number of intrinsically disordered stretches of residues defining a large region
spanning between residues 190 and 409 (Figure 4.2) (Romero et al., 2001). Although Sir3 does
not contain a structured DNA binding domain, it may undergo an induced structural change upon
binding/interacting with its cognate partner, similarly to what has been described for other
intrinsically disordered DNA binding proteins (Schulz, 1979; Weis et al., 1990; Spolar and
Record, 1994; Huth et al., 1997). The high mobility group box domain of HMG proteins
(described in Chapter 1) mediates its DNA binding through either a sequence-dependent manner
(one HMG box) or with little to no sequence specificity (requiring two or more HMG-boxes)
(Wen et al., 1989). Biophysical measurements have indicated that HMGA proteins undergo
conformational changes upon binding to their DNA targets (Huth et al., 1997). Another example
of a DNA-binding protein unstructured in the absence of DNA is the GCN4 protein, a eukaryotic
122
transcriptional activator (Hope and Struhl, 1987). This protein has also been characterized to
undergo induced local folding that generates a stable helical structure upon binding to its cognate
DNA binding site (Weis et al., 1990). Based on this work, we believe that without a defined,
structured DNA-binding domain, the intrinsic disordered segment (residues 190-409) of Sir3 is a
likely candidate to mediate its DNA-binding function.
Figure 4.1: Sir3’s Protein Primary Amino Acid Sequence. Green highlights signify amino acids with basic charges that could interact with the negatively charged backbone of nucleic
acids.
123
Figure 4.2: Degree of Intrinsic Disorder in Sir3’s Primary Structure. PONDR Score = Degree of Disorder (> 0.5 indicates disordered sequence) (Romero et al., 2001); VL-XT= Integrated predictor
of order/disorder domains; Black bars indicate large segments of intrinsic disorder
Similar to Sir3’s lack of an identified structured DNA binding domain, MeCP2 lacks a
defined structure capable of mediating histone tail binding ability. The only structured domain
described for MeCP2 is the MBD, which has classically been characterized to impart a
preference for methylated DNA (Lewis et al., 1992; Ballestar et al., 2000). Analysis of the
primary sequence of MeCP2 reveals acidic residues interspersed throughout its length. These
acidic patches could generate possible sites of interactions with the basic residues prevalent in
core histone N-termini. In addition, the MeCP2 protein contains several regions of high intrinsic
disorder (Figure 4.4) that may account for its ability to interact with the N-termini of core
histones H2A, H3, and H4 (Figure 3.3). Within these ID regions may reside the key residues that
interact with specific charges on the unstructured N-termini of H2A, H3, and H4. The inability
124
for wild type MeCP2 to bind to the H2B N-terminus (and the GST-GFP peptide) suggests that
MeCP2 is capable of recognizing specific amino acid sequences (possibly including modified
residues), and is not binding indiscriminately to short peptides. These studies do not indicate
whether this binding can induce formation of a secondary structure within MeCP2, but the
potential remains for this mechanism based on other examples of protein folding upon binding to
the biological relevant target (Bothner et al., 2001; Li et al., 2000). One such example of
binding-induced folding of disordered protein segments involves the Hdm2 (human double
minute 2) and the Arf (ADP Ribosylation Factor) proteins, known to be unstructured as
individual monomers in aqueous solutions (DiGiammariono et al., 2001). When these two
proteins form a heterodimer, they transition from a state of intrinsic disorder to a β-sheet
structure (Bothner et al., 2001). The IA3 protein, an inhibitor of yeast proteinase A (YprA), is
another example of an intrinsically disordered protein folding upon binding to its recognition site
(Dreyer et al., 1985; Green et al., 2004). This 68 aa protein was found to be unstructured when
uncomplexed, but adopts a near perfect α-helix between residues 2-32 when interacting with the
YprA’s active site cleft (Li et al., 2000). The remaining residues were accounted for as random
coil in the X-ray structure, which may indicate a persistent state of disorder in the unbound
sections of IA3 protein. As MeCP2 displays structural similarities with the above-described
proteins, it is expected to follow similar “induced folding” rules. The large amount of disorder in
the MeCP2 protein could be responsible for its ability to interact with multiple partners and play
a significant role in its multiple biological functions.
125
Figure 4.3: MeCP2’s Protein Primary Amino Acid Sequence. Red highlights signify amino acids with acidc charges that could interact with the positively charged residues of core histone N-
termini.
Figure 4.4: Degree of Intrinsic Disorder in MeCP2’s Primary Structure. PONDR Score = Degree of Disorder (> 0.5 indicates disordered sequence) (Romero et al., 2001); VL-XT= Integrated predictor
of order/disorder domains
126
CONCLUSIONS
The experiments described within this dissertation have demonstrated that the CAPs Sir3
and MeCP2 have novel chromatin interactions outside of what had originally been described.
These expanded roles may contribute to explaining their roles in multiple complex cellular
mechanisms. Each of the CAPs interacting partners may contribute to their varying cellular
functions, but the exact role and biological significance of each partner remains only partially
defined. The findings presented in this dissertation contribute to the characterization of the newly
discovered Sir3-DNA binding ability, and the identification of additional chromatin determinants
for MeCP2. Biochemical evidence also connect the binding versatility of both Sir3 and MeCP2
with their intrinsically disorder structure.
The differential binding displayed by the two RTT mutants tested in this work, supports
the theory that MeCP2’s chromatin contacts play an important biological role during
development. The loss of N-termini interaction exhibited in MeCP2 R168X truncation mutants
indicates that the TRD or CTD domains play a role in histone interaction. In addition, all three
domains correspond to regions of high intrinsic disorder (Figure 4.5) that are possible regions
linked to core histone N-termini binding. MeCP2’s interaction with histone tails also opens the
possibility for direct connections with the “histone code”. Histone PTMs could play a role in the
discriminating behavior of this CAP. MeCP2 itself can be post-translationally modified, creating
an additional level of regulation. Consequently, the PTM status both of the CAP and the histone
N-termini and the corresponding levels of local DNA methylation could all contribute to this
CAP’s proper localization and/or function (Zhou et al., 2006; Tao et al., 2009). Interestingly,
hyper-acetylation of histone N-termini did not influence MeCP2-chromatin interactions as
127
demonstrated by EMSA. It did, however impact the level of compaction (Figure 3.5, B). This
may indicate that MeCP2 can bind to chromatin before histone tail deacetylation. This is
consistent with MeCP2’s ability to associate with Sin3 and HDAC1 as it would modify the local
chromatin landscape. The newly discovered neuronal-specific hydroxymethyl-DNA may also be
involved in MeCP2 distribution and/or function, and deserves attention in the future as it may
prove to be a target for preferential binding in further studies with implications for proper
neuronal gene expression (Kriaucionis and Heintz 2009). These hydroxymethyl-DNA studies
and additional MeCP2 mutational studies will be needed to fully characterize the role of MeCP2
and chromatin interactions in neuronal gene regulation; a topic more fully explored in the Future
Studies section at the end of this dissertation.
Figure 4-5: Alignment between Sir3’s known Domains and Calculated Intrinsic Disorder Regions. PONDR Score = Degree of Disorder (> 0.5 indicates disordered sequence) (Romero et al., 2001); VL-XT= Integrated predictor
of order/disorder domains
Sir3 cooperative DNA binding is another example of how multiple chromatin
determinants may be important for proper CAP functioning. Our results demonstrate that Sir3
128
binding and its spreading over associated genomic silenced regions may occur in multiple
sequential steps (Figure 2.8). Recruitment of Sir3 would proceed through previously described
interactions with the Rap1 protein, with the level of histone-tail acetylation being an important
factor for initial chromatin interactions. Upon binding to chromatin via histone tails, subsequent
DNA-binding promotes cooperative Sir3-DNA binding possibly through local folding of Sir3’s
intrinsic disordered regions. As Sir3 does not possess a defined DNA binding motif, the
intrinsically disordered regions appear to be good candidates for mediation of Sir3-DNA
interactions. The location of Sir3 DNA binding regions was proposed to reside between residues
214 and 380 by Dr. Connelly and colleagues (Connelly et al., 2006). This area corresponds to a
highly un-structured region residing between the BAH domain and AAA domain of Sir3 (Figure
4.6). Within this large region of intrinsic disorder lies a patch of basic charges. Eight out of the
sixteen amino acids, located between residues 232-247, are either lysine (6) or arginine (2), and
this is the area most likely to mediate DNA contacts. Intrinsic disorder is commonly associated
with eukaryotic transcription factors, but Sir3 may be an example of a silencing factor using its
ID region as an additional targeting mechanism for chromatin interaction. Further in vitro and
more importantly in vivo studies will be required to determine the biological importance of Sir3’s
cooperative DNA-binding function and will be described in the following Future Studies section.
129
Figure 4.6: Alignment between Sir3’s known Domains and Calculated Intrinsic Disorder Regions. PONDR Score = Degree of Disorder (> 0.5 indicates disordered sequence) (Romero et al., 2001); VL-XT= Integrated predictor
of order/disorder domains
Link between Architectural Proteins, Histone PTM, and ATP-dependent Remodelers.
The two CAPs studied in this project, Sir3 and MeCP2, have already been the topic of
thorough investigation to characterize their interactions with chromatin. Our findings describing
and defining multiple chromatin determinants for CAP-chromatin interactions suggests that other
CAPs, in particular other architectural proteins, might benefit from similar treatment. Due to
their structural similarities, the other members of the MBD family of proteins, in particular,
should be assessed for any interactions with core histone N-termini. As with other CAPs that
interact with histone tails (ex. Sir3, HP1), the PTM status of the unstructured histone N-termini
may play a role in CAPs affinity and selectivity (Strahl and Allis 2000; Georgel et al., 2001;
Bannister et al., 2001; Lachner et al., 2001). Additional experiments will be needed to determine
130
which, if any, PTMs influence the localization and function of MeCP2 and possibly other MBD
proteins. With MeCP2 being shown to be able to condense chromatin, and two other MBD
family members, MBD2 and MBD3, found to associate with chromatin remodelers, the
possibility of a link between histone N-termini PTM signaling, MBD proteins localization, and
chromatin structure alterations is reasonably high (Georgel et al., 2003; Feng and Zhang, 2001;
Wade et al., 1999; Denslow and Wade 2007), although not described in the current accepted
mechanism. The addition of histone N-termini PTM as an active component for MeCP2
recognition would result in an increased versatility. As multiple CAPs share similar recognition
domains and mode of action, this is likely to influence their contribution to multiple cellular
processes. Identifying and characterizing the influence and function of each CAP-chromatin
interaction is critically important for the understanding of the dynamics in higher order
chromatin structure. With chromatin structure now recognized to play an essential role in
eukaryotic cell development, differentiation, and propagation, understanding these structural
interactions and their potential role in combination with any mutational errors is the first step
towards the design of corrective measures or treatments.
FUTURE STUDIES
The studies described above indicate novel chromatin interactions for Sir3 and MeCP2
which may contribute to their biological activity; in addition to the functions they have
previously been assigned. Further in vitro studies on both Sir3 and MeCP2 will be needed to
identify the region(s) responsible for these additional chromatin-associated interactions.
Subsequent in vivo experiments using these results will be required to confirm the biological
significance of these CAPs-chromatin interactions.
131
To determine the in vivo role for Sir3’s DNA binding ability, the residues necessary for
interaction must first be identified. Sir3 truncation mutants followed by subsequent point
mutations over identified areas could be used to assess DNA binding with in vitro EMSAs as
described in Chapter 2. The area spanning residues 214 and 380 and specifically the positive
residues between aa 232-247 are prime targets for this study. Once the key residues have been
identified, the effect of DNA-binding loss on Sir3 silencing could be analyzed. The contribution
of Sir3’s DNA-binding on chromatin condensation could be identified in vitro through EMSAs
with DNA-binding deficient mutants and 208-12 NA. The in vivo effect of loss of Sir3 DNA
binding can be assessed by inserting an expression vector (eg. Yep13) with the DNA-binding
deficient mutated Sir3 gene identified through EMSA into Sir3 null yeast strains (Shore et al.,
1984; Renauld et al., 1993). By increasing or decreasing the dosage of mutant Sir3 and assessing
the telomeric position effect attributed previously to Sir3 dosage (Renauld et al., 1993), the
importance of Sir3’s cooperative DNA-binding on the spread of genomic silencing could be
addressed.
Additional Sir3-chromatin interactions and the role of other potential co-regulators that
influence binding could be assessed with similar EMSA studies used in this dissertation. The O-
Acetyl-ADP-ribose, a byproduct of Sir2’s enzymatic activity, has recently been described to
influence Sir3’s chromatin binding in vitro (Liou et al., 2005; Martino et al., 2009). This
influence on binding is proposed to occur through a conformational change in Sir3’s structure
upon binding of O-Acetyl-ADP-ribose to the AAA domain. To assess if this effect is mediated
by Sir3’s DNA binding, O-Acetyl-ADP-ribose can be added at increasing concentration to the
Sir3-DNA binding reaction mix and assayed for alterations in electrophoretic mobility.
Additionally, the effect of specific histone variants could be investigated. For example, H2A.Z
132
has been characterized to reside in boundary regions of silenced chromatin (Meneghini et al.,
2003) and may influence Sir3 recruitment. This histone variant could be incorporated in place of
H2A during nucleosomal array reconstitutions, and assayed similarly by EMSA, as described
above. This would reveal whether the incorporation of the H2A.Z variant is sufficient to
influence Sir3-silencing by assessing if any change in Sir3-chromatin binding/spreading occurs.
To ascertain the effect of histone N-termini binding on MeCP2 biological function, the
regions and residues responsible for tail interactions must first be delineated. Similar to what was
described for Sir3 mutations, additional truncated mutants of MeCP2 will need to be created,
spanning the region between aa 168 and the C-terminal end. This area has been identified as
critical for MeCP2-histone N-termini interaction/s to occur (Figure 3.6). By using the GST-
histone N-termini fusion pull-down assay in combination with these generated mutants, we
would expect to identify sequences responsible for interaction. Subsequent point mutations over
these regions could then be tested to identify the key residues required for histone tail interaction.
These point mutants could then be used to generate transgenic mice and monitored for detectable
phenotypic behavior in a manner similar to what was described by Pelka and co-workers (2006).
To characterize the molecular mechanism of action, ChIP-chip and/or ChIP-seq in combination
with microarray or Q-PCR strategies could be used to monitor changes in MeCP2 localization
and corresponding effects on gene expression (Yasui et al., 2007; John, et al., 2008).
Additionally, long-distance interactions and global chromatin effects may be revealed through
analysis of MeCP2-histone binding deficient mutants in the context of formation of silent-
chromatin loop. This mechanism has been recently proposed to explain how a gene regulating
mechanism of wt MeCP2 could affect the Dlx5-Dlx6 locus in mouse (Horike et al., 2005). This
133
MeCP2-induced chromatin looping was lost in MeCP2 null mice, but the role of histone N-
termini in this silencing mechanism has not been addressed.
Other RTT mutants should also be assessed for differences in core histone N-termini
interactions. This would allow for the determination of the differential binding amongst the
common RTT mutations. The GST-histone N-termini fusion pull-down assay described above
could be employed for this purpose. Further MeCP2 studies will be needed to determine the
influence of the histone code on MeCP2 recruitment and function. A specific system has recently
been developed making use of short modified peptides to assess the effect of individual and
combinatorial PTMs on protein binding efficiency (personal communication: Denu and Garske).
This system is based on a modified Western Blot method combined with an array of various
modified peptides deposited on a nitrocellulose membrane. After incubation with appropriate
templates, the membrane is probed, analyzed by classical immunofluorescence methods and the
specificity for certain PTMs are assessed. This system would be ideally suited to analyze the
effects that histone PTMs have on wild type MeCP2 binding. Known RTT mutants could also be
tested in this manner to investigate differential binding. The biological function of subsequently
identified PTM preferences could then be confirmed in vivo by ChIP, or even ChIP-seq for
localization comparisons (John et al., 2008).
5-hydroxymethyl-DNA, a recently discovered neuronal DNA modification, could also
play a role as a key target for MeCP2-dependent gene regulation during development, as was
suggested by Kriaucionis and Heintz (2009). An in vitro competition South-Western assay of
various combinations of unmodified DNA, methylated DNA, and hydroxymethyl-DNA could be
performed with MeCP2 to analyze the preference between these various types of DNA
modifications (Ballestar et al., 2000). The enthropy of binding could also be assessed to measure
134
the strength of MeCP2 interaction with these three DNA substrates (Ghosh et al., 2008). To
determine if a correlation exists between localization of this modification and MeCP2
recruitment, ChIP-sequencing comparing pull-downs could be used to reveal if they overlap in
the genome (John et al., 2008). Additionally, the RTT mutations can be tested in a similar
manner to ascertain if a link exists between RTT and MeCP2-hydroxymethyl-DNA interaction.
Identifying the components of chromatin and how individual chromatin associated
proteins contribute to the alteration of chromatin higher order structure is essential in
understanding gene regulation. Developmental disorders such as Rett Syndrome are incredibly
difficult to treat until the complete mechanism of gene regulation and genome maintenance is
defined. Ascertaining the individual components that contribute to the formation of chromatin
higher order structure will lead to therapies to inhibit or correct deficiencies found in some
patient’s genomes and/or transcriptomes. These proposed studies will advance the understanding
of the mechanism of chromatin condensation by these two CAPs.
135
REFERENCES
Adams VH, McBryant SJ, Wade PA, Woodcock CL, Hansen JC. Intrinsic disorder and
autonomous domain function in the multifunctional nuclear protein, MeCP2. J Biol Chem. 2007 May 18;282(20):15057-64.
Agresti A, Scaffidi P, Riva A, Caiolfa VR, Bianchi ME. GR and HMGB1 interact only within
chromatin and influence each other's residence time. Mol Cell. 2005 Apr 1;18(1):109-21. Ajiro K, Yoda K, Utsumi K, Nishikawa Y. Alteration of cell cycle-dependent histone
phosphorylations by okadaic acid. Induction of mitosis-specific H3 phosphorylation and chromatin condensation in mammalian interphase cells. J Biol Chem. 1996 May 31;271(22):13197-201.
Allain FH, Gubser CC, Howe PW, Nagai K, Neuhaus D, Varani G. Specificity of
ribonucleoprotein interaction determined by RNA folding during complex formulation. Nature. 1996 Apr 18;380(6575):646-50.
Anand R, Marmorstein R. Structure and mechanism of lysine-specific demethylase enzymes. J
Biol Chem. 2007 Dec 7;282(49):35425-9. Amir RE, Van den Veyver IB, Wan M, Tran CQ, Francke U, Zoghbi HY. Rett syndrome is
caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat Genet. 1999 Oct;23(2):185-8.
Aparicio OM, Billington BL, Gottschling DE. Modifiers of position effect are shared between
telomeric and silent mating-type loci in S. cerevisiae. Cell. 1991 Sep 20;66(6):1279-87. Arkady C, et al. Histone H2AX phosphorylation is dispensable for the initial recognition of DNA
breaks. Nature Cell Biol. 2003;5:675–679. Arya G, Schilck T. Role of histone tails in chromatin folding revealed by a mesoscopic
oligonucleosome model. Proc Natl Acad Sci U S A. 2006 Oct 31;103(44):16236-41. Arya G, Schlick T. A tale of tails: how histone tails mediate chromatin compaction in different
salt and linker histone environments. J Phys Chem A. 2009 Apr 23;113(16):4045-59. Badugu R, Yoo Y, Singh PB, Kellum R. Mutations in the heterochromatin protein 1 (HP1) hinge
domain affect HP1 protein interactions and chromosomal distribution. Chromosoma. 2005 Feb;113(7):370-84.
Ballestar E, Yusufzai TM, Wolffe AP. Effects of Rett syndrome mutations of the methyl-CpG
binding domain of the transcriptional repressor MeCP2 on selectivity for association with methylated DNA. Biochemistry. 2000 Jun 20;39(24):7100-6.
136
Barton AB, Kaback DB.Telomeric silencing of an open reading frame in Saccharomyces cerevisiae.Genetics. 2006 Jun;173(2):1169-73.
Bazett-Jones DP, Côté J, Landel CC, Peterson CL, Workman JL. The SWI/SNF complex creates
loop domains in DNA and polynucleosome arrays and can disrupt DNA-histone contacts within these domains. Mol Cell Biol. 1999 Feb;19(2):1470-8.
Becker PB, Wu C. Cell-free system for assembly of transcriptionally repressed chromatin from
Drosophila embryos. Mol Cell Biol. 1992 May;12(5):2241-9. Bell SP, Mitchell J, Leber J, Kobayashi R, Stillman B. The multidomain structure of Orc1p
reveals similarity to regulators of DNA replication and transcriptional silencing. Cell. 1995 Nov 17;83(4):563-8.
Belova GI, Postnikov YV, Furusawa T, Birger Y, Bustin M. Chromosomal Protein HMGN1
Enhances the Heat Shock-induced Remodeling of Hsp70 Chromatin. J. Biol. Chem., March 28, 2008; 283(13): 8080 - 8088.
Berger SL. The complex language of chromatin regulation during transcription. Nature. 2007
May 24;447(7143):407-12. Berlingieri MT, Manfioletti G, Santoro M, Bandiera A, Visconti R, Giancotti V, Fusco A.
Inhibition of HMGI-C protein synthesis suppresses retrovirally induced neoplastic transformation of rat thyroid cells. Mol Cell Biol. 1995 Mar;15(3):1545-53.
Bernard D, Gil J, Dumont P, Rizzo S, Monté D, Quatannens B, Hudson D, Visakorpi T, Fuks F,
de Launoit Y. The methyl-CpG-binding protein MECP2 is required for prostate cancer cell growth. Oncogene. 2006 Mar 2;25(9):1358-66.
Bernstein E, Hake SB. The nucleosome: a little variation goes a long way. Biochem Cell Biol.
2006 Aug;84(4):505-17.. Berger SL. The complex language of chromatin regulation during transcription. Nature. 2007
May 24;447(7143):407-12.
Bird AP, Wolffe AP. Methylation-induced repression--belts, braces, and chromatin. Cell. 1999 Nov 24;99(5):451-4. Boisvert FM, Chénard CA, Richard S. Protein interfaces in signaling regulated by arginine
methylation. Sci STKE. 2005 Feb 15;2005(271). Bothner B, Lewis WS, DiGiammarino EL, Weber JD, Bothner SJ, Kriwacki RW. Defining the
molecular basis of Arf and Hdm2 interactions. J Mol Biol. 2001 Nov 23;314(2):263-77. Boulikas T. DNA strand breaks alter histone ADP-ribosylation. Proc Natl Acad Sci U S A. 1989
May;86(10):3499-503.
137
Boulikas T. Poly(ADP-ribosylated) histones in chromatin replication. J Biol Chem. 1990 Aug 25;265(24):14638-47.
Bourhis JM, Johansson K, Receveur-Bréchot V, Oldfield CJ, Dunker KA, Canard B, Longhi S.
The C-terminal domain of measles virus nucleoprotein belongs to the class of intrinsically disordered proteins that fold upon binding to their physiological partner. Virus Res. 2004 Feb;99(2):157-67.
Bourns BD, Alexander MK, Smith AM, Zakian VA. Sir proteins, Rif proteins, and Cdc13p bind
Saccharomyces telomeres in vivo. Mol Cell Biol. 1998 Sep;18(9):5600-8. Breiling A, Turner BM, Bianchi ME, Orlando V. General transcription factors bind promoters
repressed by Polycomb group proteins. Nature. 2001 Aug 9;412(6847):651-5. Brockdorff N., Ashworth, A., Kay, G.F., McCabe, V.M., Norris, D.P., Cooper, P.J., Swift, S.,
and Rastan, S. 1992. The product of the mouse Xist gene is a 15-kb inactive X-specific transcript containing no conserved ORF and located in the nucleus. Cell 71: 515–526.
Brown C.J., Hendrich, B.D., Rupert, J.L., Lafreniere, R.G., Xing, Y., Lawrence, J., and Willard,
H.F. 1992. The human XIST gene: Analysis of a 17-kb inactive X-specific RNA that contains conserved repeats and is highly localized within the nucleus. Cell 71: 527–542.
de Bruin D, Kantrow SM, Liberatore RA, Zakian VA. Telomere folding is required for the stable
maintenance of telomere position effects in yeast. Mol Cell Biol. 2000 Nov;20(21):7991-8000.
Buchman AR, Lue NF, Kornberg RD. Connections between transcriptional activators, silencers,
and telomeres as revealed by functional analysis of a yeast DNA-binding protein. Mol Cell Biol. 1988 Dec;8(12):5086-99.
Bühler M, Moazed D. Transcription and RNAi in heterochromatic gene silencing. Nat Struct
Mol Biol. 2007 Nov 5;14(11):1041-1048. Burzio LO, Riquelme PT, Koide SS. ADP ribosylation of rat liver nucleosomal core histones. J
Biol Chem. 1979 Apr 25;254(8):3029-37. Buschdorf JP, Strätling WH. A WW domain binding region in methyl-CpG-binding protein
MeCP2: impact on Rett syndrome. J Mol Med. 2004 Feb;82(2):135-43. Carmen AA, Milne L, Grunstein M. Acetylation of the yeast histone H4 N terminus regulates its
binding to heterochromatin protein SIR3. J Biol Chem. 2002 Feb 15;277(7):4778-81. Carney RM, Wolpert CM, Ravan SA, Shahbazian M, Ashley-Koch A, Cuccaro ML, Vance JM,
Pericak-Vance MA. Identification of MeCP2 mutations in a series of females with autistic disorder. Pediatr Neurol. 2003 Mar;28(3):205-11.
138
Cary PD, Moss T, Bradbury EM. High-resolution proton-magnetic-resonance studies of chromatin core particles. Eur J Biochem. 1978 Sep 1;89(2):475-82.
Cedar H, Bergman Y. Linking DNA methylation and histone modification: patterns and paradigms. Nat Rev Genet. 2009 May;10(5):295-304.
Chahrour M, Jung SY, Shaw C, Zhou X, Wong ST, Qin J, Zoghbi HY. MeCP2, a key
contributor to neurological disease, activates and represses transcription. Science. 2008 May
Cheng J, Kapranov P, Drenkow J, Dike S, Brubaker S, Patel S, Long J, Stern D, Tammana H,
Helt G, Sementchenko V, Piccolboni A, Bekiranov S, Bailey DK, Ganesh M, Ghosh S, Bell I, Gerhard DS, Gingeras TR. Transcriptional maps of 10 human chromosomes at 5-nucleotide resolution. Science. 2005 May 20;308(5725):1149-54. 30;320(5880):1224-9.
Cheung P, Tanner KG, Cheung WL, Sassone-Corsi P, Denu JM, Allis CD: Synergistic coupling
of histone H3 phosphorylation and acetylation in response to epidermal growth factor stimulation. Mol Cell 2000, 5:905-915.
Chiappetta G, Bandiera A, Berlingieri MT, Visconti R, Manfioletti G, Battista S, Martinez-Tello
FJ, Santoro M, Giancotti V, Fusco A. The expression of the high mobility group HMGI (Y) proteins correlates with the malignant phenotype of human thyroid neoplasias. Oncogene. 1995 Apr 6;10(7):1307-14.
Clayton AL, Rose S, Barratt MJ, Mahadevan LC: Phosphoacetylation of histone H3 on c-fos-
and c-jun-associated nucleosomes upon gene activation. EMBO J 2000, 19:3714-3726. Clemson C.M., McNeil, J.A., Willard, H.F., and Lawrence, J.B. 1996. XIST RNA paints the
inactive X chromosome at interphase: Evidence for a novel RNA involved in nuclear/chromosome structure. J. Cell Biol. 132: 259–275.
Connelly JJ, Yuan P, Hsu HC, Li Z, Xu RM, Sternglanz R. Structure and function of the
Saccharomyces cerevisiae Sir3 BAH domain. Mol Cell Biol. 2006 Apr;26(8):3256-65. Corona DF, Längst G, Clapier CR, Bonte EJ, Ferrari S, Tamkun JW, Becker PB. ISWI is an
ATP-dependent nucleosome remodeling factor. Mol Cell. 1999 Feb;3(2):239-45. Crippa MP, Alfonso PJ, Bustin M. Nucleosome core binding region of chromosomal protein
HMG-17 acts as an independent functional domain. J Mol Biol. 1992 Nov 20;228(2):442-9.
Cuthbert GL, Daujat S, Snowden AW, Erdjument-Bromage H, Hagiwara T, Yamada M,
Schneider R, Gregory PD, Tempst P, Bannister AJ, Kouzarides T. Histone deimination antagonizes arginine methylation. Cell. 2004 Sep 3;118(5):545-53.
139
Déclais AC, Fogg JM, Freeman AD, Coste F, Hadden JM, Phillips SE, Lilley DM. The complex between a four-way DNA junction and T7 endonuclease I. EMBO J. 2003 Mar 17;22(6):1398-409.
Dellino GI, Schwartz YB, Farkas G, McCabe D, Elgin SC, Pirrotta V. Polycomb silencing
blocks transcription initiation. Mol Cell. 2004 Mar 26;13(6):887-93. Denell RE. Homoeosis in Drosophila. I. Complementation studies with revertants of Nasobemia.
Genetics. 1973 Oct;75(2):279-97. Denslow SA, Wade PA. The human Mi-2/NuRD complex and gene regulation. Oncogene. 2007
Aug 13;26(37):5433-8. DiGiammarino EL, Filippov I, Weber JD, Bothner B, Kriwacki RW. Solution structure of the
p53 regulatory domain of the p19Arf tumor suppressor protein. Biochemistry. 2001 Feb 27;40(8):2379-86.
Dong F, Hansen JC, van Holde KE. DNA and protein determinants of nucleosome positioning
on sea urchin 5S rRNA gene sequences in vitro. Proc Natl Acad Sci U S A. 1990 Aug;87(15):5724-8.
Dorigo B, Schalch T, Bystricky K, Richmond TJ. Chromatin fiber folding: requirement for the
histone H4 N-terminal tail. J Mol Biol. 2003 Mar 14;327(1):85-96. Dragich JM, Kim YH, Arnold AP, Schanen NC. Differential distribution of the MeCP2 splice
variants in the postnatal mouse brain. J Comp Neurol. 2007 Apr 1;501(4):526-42. Drewell RA, Goddard CJ, Thomas JO, Surani MA. Methylation-dependent silencing at the H19
imprinting control region by MeCP2. Nucleic Acids Res. 2002 Mar 1;30(5):1139-44. Dreyer T, Valler MJ, Kay J, Charlton P, Dunn BM. The selectivity of action of the aspartic-
proteinase inhibitor IA3 from yeast (Saccharomyces cerevisiae). Biochem J. 1985 Nov 1;231(3):777-9.
ENCODE Project Consortium, Birney E, Stamatoyannopoulos JA, Dutta A, Guigó R, Gingeras
TR, Margulies EH, Weng Z, Snyder M, Dermitzakis ET, Thurman RE, Kuehn MS, Taylor CM, Neph S, Koch CM, Asthana S, Malhotra A, Adzhubei I, Greenbaum JA, Andrews RM, Flicek P, Boyle PJ, Cao H, Carter NP, Clelland GK, Davis S, Day N, Dhami P, Dillon SC, Dorschner MO, Fiegler H, Giresi PG, Goldy J, Hawrylycz M, Haydock A, Humbert R, James KD, Johnson BE, Johnson EM, Frum TT, Rosenzweig ER, Karnani N, Lee K, Lefebvre GC, Navas PA, Neri F, Parker SC, Sabo PJ, Sandstrom R, Shafer A, Vetrie D, Weaver M, Wilcox S, Yu M, Collins FS, Dekker J, Lieb JD, Tullius TD, Crawford GE, Sunyaev S, Noble WS, Dunham I, Denoeud F, Reymond A, Kapranov P, Rozowsky J, Zheng D, Castelo R, Frankish A, Harrow J, Ghosh S, Sandelin A, Hofacker IL, Baertsch R, Keefe D, Dike S, Cheng J, Hirsch HA, Sekinger EA, Lagarde J, Abril JF, Shahab A, Flamm C, Fried C, Hackermüller J, Hertel J, Lindemeyer
140
M, Missal K, Tanzer A, Washietl S, Korbel J, Emanuelsson O, Pedersen JS, Holroyd N, Taylor R, Swarbreck D, Matthews N, Dickson MC, Thomas DJ, Weirauch MT, Gilbert J, Drenkow J, Bell I, Zhao X, Srinivasan KG, Sung WK, Ooi HS, Chiu KP, Foissac S, Alioto T, Brent M, Pachter L, Tress ML, Valencia A, Choo SW, Choo CY, Ucla C, Manzano C, Wyss C, Cheung E, Clark TG, Brown JB, Ganesh M, Patel S, Tammana H, Chrast J, Henrichsen CN, Kai C, Kawai J, Nagalakshmi U, Wu J, Lian Z, Lian J, Newburger P, Zhang X, Bickel P, Mattick JS, Carninci P, Hayashizaki Y, Weissman S, Hubbard T, Myers RM, Rogers J, Stadler PF, Lowe TM, Wei CL, Ruan Y, Struhl K, Gerstein M, Antonarakis SE, Fu Y, Green ED, Karaöz U, Siepel A, Taylor J, Liefer LA, Wetterstrand KA, Good PJ, Feingold EA, Guyer MS, Cooper GM, Asimenos G, Dewey CN, Hou M, Nikolaev S, Montoya-Burgos JI, Löytynoja A, Whelan S, Pardi F, Massingham T, Huang H, Zhang NR, Holmes I, Mullikin JC, Ureta-Vidal A, Paten B, Seringhaus M, Church D, Rosenbloom K, Kent WJ, Stone EA; NISC Comparative Sequencing Program; Baylor College of Medicine Human Genome Sequencing Center; Washington University Genome Sequencing Center; Broad Institute; Children's Hospital Oakland Research Institute, Batzoglou S, Goldman N, Hardison RC, Haussler D, Miller W, Sidow A, Trinklein ND, Zhang ZD, Barrera L, Stuart R, King DC, Ameur A, Enroth S, Bieda MC, Kim J, Bhinge AA, Jiang N, Liu J, Yao F, Vega VB, Lee CW, Ng P, Shahab A, Yang A, Moqtaderi Z, Zhu Z, Xu X, Squazzo S, Oberley MJ, Inman D, Singer MA, Richmond TA, Munn KJ, Rada-Iglesias A, Wallerman O, Komorowski J, Fowler JC, Couttet P, Bruce AW, Dovey OM, Ellis PD, Langford CF, Nix DA, Euskirchen G, Hartman S, Urban AE, Kraus P, Van Calcar S, Heintzman N, Kim TH, Wang K, Qu C, Hon G, Luna R, Glass CK, Rosenfeld MG, Aldred SF, Cooper SJ, Halees A, Lin JM, Shulha HP, Zhang X, Xu M, Haidar JN, Yu Y, Ruan Y, Iyer VR, Green RD, Wadelius C, Farnham PJ, Ren B, Harte RA, Hinrichs AS, Trumbower H, Clawson H, Hillman-Jackson J, Zweig AS, Smith K, Thakkapallayil A, Barber G, Kuhn RM, Karolchik D, Armengol L, Bird CP, de Bakker PI, Kern AD, Lopez-Bigas N, Martin JD, Stranger BE, Woodroffe A, Davydov E, Dimas A, Eyras E, Hallgrímsdóttir IB, Huppert J, Zody MC, Abecasis GR, Estivill X, Bouffard GG, Guan X, Hansen NF, Idol JR, Maduro VV, Maskeri B, McDowell JC, Park M, Thomas PJ, Young AC, Blakesley RW, Muzny DM, Sodergren E, Wheeler DA, Worley KC, Jiang H, Weinstock GM, Gibbs RA, Graves T, Fulton R, Mardis ER, Wilson RK, Clamp M, Cuff J, Gnerre S, Jaffe DB, Chang JL, Lindblad-Toh K, Lander ES, Koriabine M, Nefedov M, Osoegawa K, Yoshinaga Y, Zhu B, de Jong PJ. Identification and analysis of functional elements in 1% of the human genome by the ENCODE pilot project. Nature 2007; 447:799-816.
Eissenberg JC, James TC, Foster-Hartnett DM, Hartnett T, Ngan V, Elgin SC. Mutation in a
heterochromatin-specific chromosomal protein is associated with suppression of position-effect variegation in Drosophila melanogaster. Proc Natl Acad Sci U S A. 1990 Dec;87(24):9923-7.
El-Osta A, Kantharidis P, Zalcberg JR, Wolffe AP. Precipitous release of methyl-CpG binding
protein 2 and histone deacetylase 1 from the methylated human multidrug resistance gene (MDR1) on activation. Mol Cell Biol. 2002 Mar;22(6):1844-57.
141
Erzberger JP, Berger JM. Evolutionary relationships and structural mechanisms of AAA+ proteins. Annu Rev Biophys Biomol Struct. 2006;35:93-114.
Fan JY, Rangasamy D, Luger K, Tremethick DJ. H2A.Z alters the nucleosome surface to
promote HP1alpha-mediated chromatin fiber folding. Mol Cell. 2004 Nov 19;16(4):655-61.
Feng Q, Zhang Y. The MeCP1 complex represses transcription through preferential binding,
remodeling, and deacetylating methylated nucleosomes. Genes Dev. 2001 Apr 1;15(7):827-32.
Fernandez-Capetillo O, Allis CD, Nussenzweig A. Phosphorylation of histone H2B at DNA
double-strand breaks. J Exp Med. 2004 Jun 21;199(12):1671-7. Ferrari S, Harley VR, Pontiggia A, Goodfellow PN, Lovell-Badge R, Bianchi ME. SRY, like
HMG1, recognizes sharp angles in DNA. EMBO J. 1992 Dec;11(12):4497-506. Fire A, Xu S, Montgomery MK, Kostas SA, Driver SE, Mello CC. Potent and specific genetic
interference by double-stranded RNA in Caenorhabditis elegans. Nature. 1998 Feb 19;391(6669):806-11.
Fletcher TM, Serwer P, Hansen JC. Quantitative analysis of macromolecular conformational
changes using agarose gel electrophoresis: application to chromatin folding. Biochemistry. 1994 Sep 13;33(36):10859-63.
Fletcher TM, Hansen JC. Core histone tail domains mediate oligonucleosome folding and
nucleosomal DNA organization through distinct molecular mechanisms. J Biol Chem. 1995 Oct 27;270(43):25359-62.
Francis NJ, Kingston RE, Woodcock CL. Chromatin compaction by a polycomb group protein
complex. Science. 2004 Nov 26;306(5701):1574-7. Frechette, A., A. Huletsky, R. J. Aubin, G. de Murcia, P. Mandel, A. Lord, G. Grondin, and G.
G. Poirier. 1985. Poly(ADP-ribosyl)ation of chromatin: kinetics of relaxation and its effect on chromatin solubility. Can. J. Biochem. Cell Biol. 63:764–773.
Free A, Wakefield RI, Smith BO, Dryden DT, Barlow PN, Bird AP. DNA recognition by the
methyl-CpG binding domain of MeCP2. J Biol Chem. 2001 Feb 2;276(5):3353-60. Fukagawa T, Nogami M, Yoshikawa M, Ikeno M, Okazaki T, Takami Y, Nakayama T,
Oshimura M. Dicer is essential for formation of the heterochromatin structure in vertebrate cells. Nat Cell Biol. 2004 Aug;6(8):784-91.
Fuks F, Hurd PJ, Wolf D, Nan X, Bird AP, Kouzarides T. The methyl-CpG-binding protein
MeCP2 links DNA methylation to histone methylation. J Biol Chem. 2003 Feb 7;278(6):4035-40.
142
Georgel P, Demeler B, Terpening C, Paule MR, van Holde KE. Binding of the RNA polymerase
I transcription complex to its promoter can modify positioning of downstream nucleosomes assembled in vitro. J Biol Chem. 1993 Jan 25;268(3):1947-54.
Georgel PT, Tsukiyama T, Wu C. Role of histone tails in nucleosome remodeling by Drosophila
NURF. EMBO J. 1997 Aug 1;16(15):4717-26. Georgel PT, Palacios DeBeer MA, Pietz G, Fox CA, Hansen JC. Sir3-dependent assembly of
supramolecular chromatin structures in vitro.Proc Natl Acad Sci U S A. 2001 Jul 17;98(15):8584-9.
Georgel PT, Robert CH. Differential core histone binding behavior: RNA polymerase I promoter
region vs 5S rDNA positioning DNA sequences. Cell Biochem Biophys. 2002;37(1):1-13.
Georgel PT, Fletcher TM, Hager GL, Hansen JC. Formation of higher-order secondary and
tertiary chromatin structures by genomic mouse mammary tumor virus promoters. Genes Dev. 2003 Jul 1;17(13):1617-29.
Georgel PT, Horowitz-Scherer RA, Adkins N, Woodcock CL, Wade PA, Hansen JC. Chromatin
compaction by human MeCP2. Assembly of novel secondary chromatin structures in the absence of DNA methylation. J Biol Chem. 2003 Aug 22;278(34):32181-8.
Gerlitz G, Hock R, Ueda T, Bustin M. The dynamics of HMG protein-chromatin interactions in
living cells. Biochem Cell Biol. 2009 Feb;87(1):127-37. Ghosh RP, Horowitz-Scherer RA, Nikitina T, Gierasch LM, Woodcock CL. Rett syndrome-
causing mutations in human MeCP2 result in diverse structural changes that impact folding and DNA interactions. J Biol Chem. 2008 Jul 18;283(29):20523-34.
Gonzalez-Zulueta M, Bender CM, Yang AS, Nguyen T, Beart RW, Van Tornout JM, Jones PA.
Methylation of the 5' CpG island of the p16/CDKN2 tumor suppressor gene in normal and transformed human tissues correlates with gene silencing.Cancer Res. 1995 Oct 15;55(20):4531-5.
Gordon F, Luger K, Hansen JC. The Core Histone N-terminal Tail Domains Function
Independently and Additively during Salt-dependent Oligomerization of Nucleosomal Arrays. J Biol Chem. 2005 Oct 7;280(40):33701-6.
Gottesfeld JM, Butler PJ. Structure of transcriptionally-active chromatin subunits. Nucleic Acids
Res. 1977 Sep;4(9):3155-73. Gotta M, Palladino F, Gasser SM. Functional characterization of the N terminus of Sir3p.Mol
Cell Biol. 1998 Oct;18(10):6110-20.
143
Green TB, Ganesh O, Perry K, Smith L, Phylip LH, Logan TM, Hagen SJ, Dunn BM, Edison AS. IA3, an aspartic proteinase inhibitor from Saccharomyces cerevisiae, is intrinsically unstructured in solution. Biochemistry. 2004 Apr 13;43(14):4071-81.
Grigoryev SA, Woodcock CL. Chromatin structure in granulocytes. A link between tight
compaction and accumulation of a heterochromatin-associated protein (MENT). J Biol Chem. 1998 Jan 30;273(5):3082-9.
Grigoryev SA. Higher-order folding of heterochromatin: protein bridges span the nucleosome
arrays. Biochem Cell Biol. 2001;79(3):227-41. Guenther MG, Levine SS, Boyer LA, Jaenisch R, Young RA. A chromatin landmark and
transcription initiation at most promoters in human cells. Cell. 2007 Jul 13;130(1):77-88. Guo XW, Th'ng JP, Swank RA, Anderson HJ, Tudan C, Bradbury EM, Roberge M.
Chromosome condensation induced by fostriecin does not require p34cdc2 kinase activity and histone H1 hyperphosphorylation, but is associated with enhanced histone H2A and H3 phosphorylation. EMBO J. 1995 Mar 1;14(5):976-85.
Hagberg B, Aicardi J, Dias K, Ramos O. A progressive syndrome of autism, dementia, ataxia,
and loss of purposeful hand use in girls: Rett's syndrome: report of 35 cases. Ann Neurol. 1983 Oct;14(4):471-9.
Hagberg, B. Rett’s syndrome: prevalence and impact on progressive severe mental retardation in
girls. Acta Paediatr. Scand. 74, 405–408 (1985). Hammer S, Dorrani N, Dragich J, Kudo S, Schanen C. The phenotypic consequences of MECP2
mutations extend beyond Rett syndrome. Ment Retard Dev Disabil Res Rev. 2002;8(2):94-8.
Hansen JC, Ausio J, Stanik VH, van Holde KE. Homogeneous reconstituted oligonucleosomes,
evidence for salt-dependent folding in the absence of histone H1. Biochemistry. 1989 Nov 14;28(23):9129-36.
Hansen JC, Lohr D. Assembly and structural properties of subsaturated chromatin arrays. J Biol
Chem. 1993 Mar 15;268(8):5840-8. Hansen JC. Conformational dynamics of the chromatin fiber in solution: determinants,
mechanisms, and functions. Annu Rev Biophys Biomol Struct. 2002;31:361-92. Epub 2001 Oct 25.
Hassa PO, and Hottiger MO. An epigenetic code for DNA damage repair pathways? Biochem.
Cell Biol. 2005 83:270–285.
144
Hassa PO, Haenni SS, Elser M, Hottiger MO. Nuclear ADP-ribosylation reactions in mammalian cells: where are we today and where are we going? Microbiol Mol Biol Rev. 2006 Sep;70(3):789-829.
Hecht A, Laroche T, Strahl-Bolsinger S, Gasser SM, Grunstein M. Histone H3 and H4 N-termini
interact with SIR3 and SIR4 proteins: a molecular model for the formation of heterochromatin in yeast. Cell. 1995 Feb 24;80(4):583-92.
Hecht A, Strahl-Bolsinger S, Grunstein M. Spreading of transcriptional repressor SIR3 from
telomeric heterochromatin. Nature. 1996 Sep 5;383(6595):92-6. Hediger F, Gasser SM. Heterochromatin protein 1: don't judge the book by its cover! Curr Opin
Genet Dev. 2006 Apr;16(2):143-50. Epub 2006 Feb 28. Hendrich B, Bird A. Identification and characterization of a family of mammalian methyl-CpG
binding proteins. Mol Cell Biol. 1998 Nov;18(11):6538-47. Hendrich B, Hardeland U, Ng HH, Jiricny J, Bird A. The thymine glycosylase MBD4 can bind
to the product of deamination at methylated CpG sites. Nature. 1999 Sep 16;401(6750):301-4.
Hendrich B, Guy J, Ramsahoye B, Wilson VA, Bird A. Closely related proteins MBD2 and
MBD3 play distinctive but interacting roles in mouse development. Genes Dev. 2001 Mar 15;15(6):710-23.
Hill DA, Reeves R. Competition between HMG-I(Y), HMG-1 and histone H1 on four-way
junction DNA. Nucleic Acids Res. 1997 Sep 1;25(17):3523-31. Hirschhorn JN, Brown SA, Clark CD, Winston F. Evidence that SNF2/SWI2 and SNF5 activate
transcription in yeast by altering chromatin structure. Genes Dev. 1992 Dec;6(12A):2288-98.
Hite KC, Adams VH, Hansen JC. Recent advances in MeCP2 structure and function. Biochem
Cell Biol. 2009 Feb;87(1):219-27. Holliday R, Pugh JE. DNA modification mechanisms and gene activity during development.
Science. 1975 Jan 24;187(4173):226-32. Hope IA, Struhl K. GCN4, a eukaryotic transcriptional activator protein, binds as a dimer to
target DNA. EMBO J. 1987 Sep;6(9):2781-4. Hoppe GJ, Tanny JC, Rudner AD, Gerber SA, Danaie S, Gygi SP, Moazed D. Steps in assembly
of silent chromatin in yeast: Sir3-independent binding of a Sir2/Sir4 complex to silencers and role for Sir2-dependent deacetylation. Mol Cell Biol. 2002 Jun;22(12):4167-80.
145
Horike S, Cai S, Miyano M, Cheng JF, Kohwi-Shigematsu T. Loss of silent-chromatin looping and impaired imprinting of DLX5 in Rett syndrome. Nat Genet. 2005 Jan;37(1):31-40.
Hou Z, Bernstein DA, Fox CA, Keck JL. Structural basis of the Sir1-origin recognition complex
interaction in transcriptional silencing. Proc Natl Acad Sci U S A. 2005 Jun 14;102(24):8489-94.
Huertas D, Cortés A, Casanova J, Azorín F. Drosophila DDP1, a multi-KH-domain protein,
contributes to centromeric silencing and chromosome segregation. Curr Biol. 2004 Sep 21;14(18):1611-20.
Hunter T. The age of crosstalk: phosphorylation, ubiquitination, and beyond. Mol Cell. 2007 Dec
14;28(5):730-8. Huth JR, Bewley CA, Nissen MS, Evans JN, Reeves R, Gronenborn AM, Clore GM. The
solution structure of an HMG-I(Y)-DNA complex defines a new architectural minor groove binding motif. Nat Struct Biol. 1997 Aug;4(8):657-65.
Iberg AN, Espejo A, Cheng D, Kim D, Michaud-Levesque J, Richard S, Bedford MT. Arginine
methylation of the histone H3 tail impedes effector binding. J Biol Chem. 2008 Feb 8;283(6):3006-10.
Ikura T, Tashiro S, Kakino A, Shima H, Jacob N, Amunugama R, Yoder K, Izumi S, Kuraoka I,
Tanaka K, Kimura H, Ikura M, Nishikubo S, Ito T, Muto A, Miyagawa K, Takeda S, Fishel R, Igarashi K, Kamiya K. DNA damage-dependent acetylation and ubiquitination of H2AX enhances chromatin dynamics. Mol Cell Biol. 2007 Oct;27(20):7028-40.
Imbalzano AN, Kwon H, Green MR, Kingston RE. Facilitated binding of TATA-binding protein
to nucleosomal DNA. Nature. 1994 Aug 11;370(6489):481-5. Imai S, Johnson FB, Marciniak RA, McVey M, Park PU, Guarente L.Sir2: an NAD-dependent
histone deacetylase that connects chromatin silencing, metabolism, and aging.Cold Spring Harb Symp Quant Biol. 2000;65:297-302.
Irving JA, Shushanov SS, Pike RN, Popova EY, Brömme D, Coetzer TH, Bottomley SP,
Boulynko IA, Grigoryev SA, Whisstock JC. Inhibitory activity of a heterochromatin-associated serpin (MENT) against papain-like cysteine proteinases affects chromatin structure and blocks cell proliferation. J Biol Chem. 2002 Apr 12;277(15):13192-201.
Ishibashi T, Thambirajah AA, Ausió J. MeCP2 preferentially binds to methylated linker DNA in
the absence of the terminal tail of histone H3 and independently of histone acetylation. FEBS Lett. 2008 Apr 2;582(7):1157-62.
Jacobs JJ, Kieboom K, Marino S, DePinho RA, van Lohuizen M. The oncogene and Polycomb-
group gene bmi-1 regulates cell proliferation and senescence through the ink4a locus. Nature. 1999 Jan 14;397(6715):164-8.
146
James TC, Elgin SC. Identification of a nonhistone chromosomal protein associated with
heterochromatin in Drosophila melanogaster and its gene. Mol Cell Biol. 1986 Nov;6(11):3862-72.
James TC, Eissenberg JC, Craig C, Dietrich V, Hobson A, Elgin SC. Distribution patterns of
HP1, a heterochromatin-associated nonhistone chromosomal protein of Drosophila. Eur J Cell Biol. 1989 Oct;50(1):170-80.
John S, Sabo PJ, Johnson TA, Sung MH, Biddie SC, Lightman SL, Voss TC, Davis SR, Meltzer
PS, Stamatoyannopoulos JA, Hager GL. Interaction of the glucocorticoid receptor with the chromatin landscape. Mol Cell. 2008 Mar 14;29(5):611-24.
Johnson LM, Kayne PS, Kahn ES, Grunstein M. Genetic evidence for an interaction between
SIR3 and histone H4 in the repression of the silent mating loci in Saccharomyces cerevisiae. Proc Natl Acad Sci U S A. 1990 Aug;87(16):6286-90.
Jones PL, Veenstra GJ, Wade PA, Vermaak D, Kass SU, Landsberger N, Strouboulis J, Wolffe
AP. Methylated DNA and MeCP2 recruit histone deacetylase to repress transcription. Nat Genet. 1998 Jun;19(2):187-91.
Kaji K, Caballero IM, MacLeod R, Nichols J, Wilson VA, Hendrich B. The NuRD component
Mbd3 is required for pluripotency of embryonic stem cells. Nat Cell Biol. 2006 Mar;8(3):285-92.
Kan PY, Lu X, Hansen JC, Hayes JJ. The H3 tail domain participates in multiple interactions
during folding and self-association of nucleosome arrays. Mol Cell Biol. 2007 Mar;27(6):2084-91.
Kan PY, Caterino TL, Hayes JJ. The H4 tail domain participates in intra- and internucleosome
interactions with protein and DNA during folding and oligomerization of nucleosome arrays. Mol Cell Biol. 2009 Jan;29(2):538-46.
Kanno M, Hasegawa M, Ishida A, Isono K, Taniguchi M. mel-18, a Polycomb group-related
mammalian gene, encodes a transcriptional negative regulator with tumor suppressive activity. EMBO J. 1995 Nov 15;14(22):5672-8.
Kaplan LJ, Bauer R, Morrison E, Langan TA, Fasman GD. The structure of chromatin
reconstituted with phosphorylated H1. Circular dichroism and thermal denaturation studies. J Biol Chem. 1984 Jul 25;259(14):8777-85.
Katayama S, Tomaru Y, Kasukawa T, Waki K, Nakanishi M, Nakamura M, Nishida H, Yap CC,
Suzuki M, Kawai J, Suzuki H, Carninci P, Hayashizaki Y, Wells C, Frith M, Ravasi T, Pang KC, Hallinan J, Mattick J, Hume DA, Lipovich L, Batalov S, Engström PG, Mizuno Y, Faghihi MA, Sandelin A, Chalk AM, Mottagui-Tabar S, Liang Z, Lenhard B, Wahlestedt C; RIKEN Genome Exploration Research Group; Genome Science Group
147
(Genome Network Project Core Group); FANTOM Consortium. Antisense transcription in the mammalian transcriptome. Science 2005;309:1564-1566.
King DA, Hall BE, Iwamoto MA, Win KZ, Chang JF, Ellenberger T. Domain structure and
protein interactions of the silent information regulator Sir3 revealed by screening a nested deletion library of protein fragments. J Biol Chem. 2006 Jul 21;281(29):20107-19.
Klein F, Laroche T, Cardenas ME, Hofmann JF, Schweizer D, Gasser SM. Localization of RAP1
and topoisomerase II in nuclei and meiotic chromosomes of yeast. J Cell Biol. 1992 Jun;117(5):935-48.
Koike N, Maita H, Taira T, Ariga H, Iguchi-Ariga SM. Identification of heterochromatin protein
1 (HP1) as a phosphorylation target by Pim-1 kinase and the effect of phosphorylation on the transcriptional repression function of HP1(1). FEBS Lett. 2000 Feb 4;467(1):17-21.
Kondo E, Gu Z, Horii A, Fukushige S. The thymine DNA glycosylase MBD4 represses
transcription and is associated with methylated p16(INK4a) and hMLH1 genes. Mol Cell Biol. 2005 Jun;25(11):4388-96.
Konkel, L. M., Enomoto, S., Chamberlain, E. M., McCune-Zierath, P., Iyadurai, S. J., and
Berman, J. A class of single-stranded telomeric DNA-binding proteins required for Rap1p localization in yeast nuclei. Proc. Natl. Acad. Sci. U.S.A 1995 92, 5558-5562.
Kreimeyer A, Wielckens K, Adamietz P, Hilz H. DNA repair-associated ADP-ribosylation in vivo. Modification of histone H1 differs from that of the principal acceptor proteins. J Biol Chem. 1984 Jan 25;259(2):890-6.
Kriaucionis S, Bird A. The major form of MeCP2 has a novel N-terminus generated by
alternative splicing. Nucleic Acids Res. 2004 Mar 19;32(5):1818-23. Kriaucionis S, Heintz N. The nuclear DNA base 5-hydroxymethylcytosine is present in Purkinje
neurons and the brain. Science. 2009 May 15;324(5929):929-30. Kristjuhan A, Wittschieben BO, Walker J, Roberts D, Cairns BR, Svejstrup JQ. Spreading of
Sir3 protein in cells with severe histone H3 hypoacetylation. Proc Natl Acad Sci U S A. 2003 Jun 24;100(13):7551-6.
Kwon H, Imbalzano AN, Khavari PA, Kingston RE, Green MR. Nucleosome disruption and
enhancement of activator binding by a human SW1/SNF complex. Nature. 1994 Aug 11;370(6489):477-81.
Kwon SH, Workman JL. The heterochromatin protein 1 (HP1) family: put away a bias toward
HP1. Mol Cells. 2008 Sep 30;26(3):217-27.
148
Kyrion G, Boakye KA, Lustig AJ. C-terminal truncation of RAP1 results in the deregulation of telomere size, stability, and function in Saccharomyces cerevisiae. Mol Cell Biol. 1992 Nov;12(11):5159-73.
Lachner M, O'Carroll D, Rea S, Mechtler K, Jenuwein T. Methylation of histone H3 lysine 9
creates a binding site for HP1 proteins. Nature. 2001 Mar 1;410(6824):116-20. Landsman D, Bustin M. Assessment of the transcriptional activation potential of the HMG
chromosomal proteins. Mol Cell Biol. 1991 Sep;11(9):4483-9. Laurenson P, Rine J.Silencers, silencing, and heritable transcriptional states. Microbiol Rev.
1992 Dec;56(4):543-60. Lee KB, Thomas JO. The effect of the acidic tail on the DNA-binding properties of the HMG1,2
class of proteins: insights from tail switching and tail removal. J Mol Biol. 2000 Nov 24;304(2):135-49.
Lee SS, Wan M, Francke U. Spectrum of MECP2 mutations in Rett syndrome. Brain Dev. 2001
Dec;23 Suppl 1:S138-43. Lewis EB. A gene complex controlling segmentation in Drosophila. Nature. 1978 Dec
7;276(5688):565-70. Lewis JD, Meehan RR, Henzel WJ, Maurer-Fogy I, Jeppesen P, Klein F, Bird A. Purification,
sequence, and cellular localization of a novel chromosomal protein that binds to methylated DNA. Cell. 1992 Jun 12;69(6):905-14.
Li M, Phylip LH, Lees WE, Winther JR, Dunn BM, Wlodawer A, Kay J, Gustchina A. The
aspartic proteinase from Saccharomyces cerevisiae folds its own inhibitor into a helix. Nat Struct Biol. 2000 Feb;7(2):113-7.
Liaw H, Lustig AJ. Sir3 C-terminal domain involvement in the initiation and spreading of
heterochromatin. Mol Cell Biol. 2006 Oct;26(20):7616-31. Liberati C, Sgarra R, Manfioletti G, Mantovani R. DNA binding of NF-Y: the effect of HMGI
proteins depends upon the CCAAT box. FEBS Lett. 1998 Aug 14;433(1-2):174-8. Lilley DM. Structures of helical junctions in nucleic acids. Q Rev Biophys. 2000 May;33(2):109-
59. Lin JJ, Zakian VA. Isolation and characterization of two Saccharomyces cerevisiae genes that
encode proteins that bind to (TG1-3)n single strand telomeric DNA in vitro. Nucleic Acids Res. 1994 Nov 25;22(23):4906-13.
149
Liou GG, Tanny JC, Kruger RG, Walz T, Moazed D. Assembly of the SIR complex and its regulation by O-acetyl-ADP-ribose, a product of NAD-dependent histone deacetylation. Cell. 2005 May 20;121(4):515-27.
Liu J, Perumal NB, Oldfield CJ, Su EW, Uversky VN, Dunker AK. Intrinsic disorder in
transcription factors. Biochemistry. 2006 Jun 6;45(22):6873-88. Lim JH, West KL, Rubinstein Y, Bergel M, Postnikov YV, Bustin M. Chromosomal protein
HMGN1 enhances the acetylation of lysine 14 in histone H3. EMBO J. 2005 Sep 7;24(17):3038-48.
Lin R, Cook RG, Allis CD. Proteolytic removal of core histone amino termini and
dephosphorylation of histone H1 correlate with the formation of condensed chromatin and transcriptional silencing during Tetrahymena macronuclear development. Genes Dev. 1991 Sep;5(9):1601-10.
Linderstrøm-Lang KU and Schellman JA. Protein structure and enzyme activit Protein structure
and enzyme activit. In The Enzymes. 1959. Boyer PD, Lardy H, Myrback K (eds). Academic Press: New York 443-510.
Lo WS, Trievel RC, Rojas JR, Duggan L, Hsu JY, Allis CD, Marmorstein R, Berger SL:
Phosphorylation of serine 10 in histone H3 is functionally linked in vitro and in vivo to Gcn5-mediated acetylation at lysine 14. Mol Cell 2000, 5:917-926.
Lohr D, Corden J, Tatchell K, Kovacic RT, Van Holde KE. Comparative subunit structure of
HeLa, yeast, and chicken erythrocyte chromatin. Proc Natl Acad Sci U S A. 1977 Jan;74(1):79-83.
Loo S, Rine J. Silencers and domains of generalized repression. Science. 1994 Jun
17;264(5166):1768-71. Luger K, Hansen JC. Nucleosome and chromatin fiber dynamics. Curr Opin Struct Biol. 2005
Apr;15(2):188-96. Luger K, Mäder AW, Richmond RK, Sargent DF, Richmond TJ.Crystal structure of the
nucleosome core particle at 2.8 A resolution.Nature. 1997 Sep 18;389(6648):251-60. Lynch SA, Whatley SD, Ramesh V, Sinha S, Ravine D. Sporadic case of fatal encephalopathy
with neonatal onset associated with a T158M missense mutation in MECP2. Arch Dis Child Fetal Neonatal Ed. 2003 May;88(3):F250-2.
Madigan JP, Chotkowski HL, Glaser RL. DNA double-strand break-induced phosphorylation of
Drosophila histone variant H2Av helps prevent radiation-induced apoptosis. Nucleic Acids Res. 2002 Sep 1;30(17):3698-705.
150
Markus MA, Hinck AP, Huang S, Draper DE, Torchia DA. High resolution solution structure of ribosomal protein L11-C76, a helical protein with a flexible loop that becomes structured upon binding to RNA. Nat Struct Biol. 1997 Jan;4(1):70-7.
Martino F, Kueng S, Robinson P, Tsai-Pflugfelder M, van Leeuwen F, Ziegler M, Cubizolles F,
Cockell MM, Rhodes D, Gasser SM. Reconstitution of yeast silent chromatin: multiple contact sites and O-AADPR binding load SIR complexes onto nucleosomes in vitro. Mol Cell. 2009 Feb 13;33(3):323-34.
McBryant SJ, Adams VH, Hansen JC. Chromatin architectural proteins. Chromosome Res.
2006;14(1):39-51. McBryant SJ, Krause C, Hansen JC. Domain organization and quaternary structure of the
Saccharomyces cerevisiae silent information regulator 3 protein, Sir3p.Biochemistry. 2006 Dec 26;45(51):15941-8.
McBryant SJ, Krause C, Woodcock CL, Hansen JC. The silent information regulator 3 protein,
SIR3p, binds to chromatin fibers and assembles a hypercondensed chromatin architecture in the presence of salt. Mol Cell Biol. 2008 Jun;28(11):3563-72.
Meehan RR, Lewis JD, Bird AP. Characterization of MeCP2, a vertebrate DNA binding protein
with affinity for methylated DNA. Nucleic Acids Res. 1992 Oct 11;20(19):5085-92. Meloni I, Bruttini M, Longo I, Mari F, Rizzolio F, D'Adamo P, Denvriendt K, Fryns JP, Toniolo
D, Renieri A. A mutation in the rett syndrome gene, MECP2, causes X-linked mental retardation and progressive spasticity in males. Am J Hum Genet. 2000 Oct;67(4):982-5.
Meneghini MD, Wu M, Madhani HD. Conserved histone variant H2A.Z protects euchromatin
from the ectopic spread of silent heterochromatin. Cell. 2003 Mar 7;112(5):725-36. Merlo A, Herman JG, Mao L, Lee DJ, Gabrielson E, Burger PC, Baylin SB, Sidransky D.
5' CpG island methylation is associated with transcriptional silencing of the tumour suppressor p16/CDKN2/MTS1 in human cancers.Nat Med. 1995 Jul;1(7):686-92.
Messmer S, Franke A, Paro R. Analysis of the functional role of the Polycomb chromo domain
in Drosophila melanogaster. Genes Dev. 1992 Jul;6(7):1241-54. Mette MF, Aufsatz W, van der Winden J, Matzke MA, Matzke AJ. Transcriptional silencing and
promoter methylation triggered by double-stranded RNA. EMBO J. 2000 Oct 2;19(19):5194-201.
Millar CB, Guy J, Sansom OJ, Selfridge J, MacDougall E, Hendrich B, Keightley PD, Bishop
SM, Clarke AR, Bird A. Enhanced CpG mutability and tumorigenesis in MBD4-deficient mice. Science. 2002 Jul 19;297(5580):403-5.
151
Miltenberger-Miltenyi G, Laccone F. Mutations and polymorphisms in the human methyl CpG-binding protein MECP2. Hum Mutat. 2003 Aug;22(2):107-15.
Minc E, Allory Y, Worman HJ, Courvalin JC, Buendia B. Localization and phosphorylation of
HP1 proteins during the cell cycle in mammalian cells. Chromosoma. 1999 Aug;108(4):220-34.
Mnatzakanian GN, Lohi H, Munteanu I, Alfred SE, Yamada T, MacLeod PJ, Jones JR, Scherer
SW, Schanen NC, Friez MJ, Vincent JB, Minassian BA. A previously unidentified MECP2 open reading frame defines a new protein isoform relevant to Rett syndrome. Nat Genet. 2004 Apr;36(4):339-41.
Moazed D, Kistler A, Axelrod A, Rine J, Johnson AD. Silent information regulator protein
complexes in Saccharomyces cerevisiae: a SIR2/SIR4 complex and evidence for a regulatory domain in SIR4 that inhibits its interaction with SIR3. Proc Natl Acad Sci U S A. 1997 Mar 18;94(6):2186-91.
Molofsky AV, Pardal R, Iwashita T, Park IK, Clarke MF, Morrison SJ.Bmi-1 dependence
distinguishes neural stem cell self-renewal from progenitor proliferation. Nature. 2003 Oct 30;425(6961):962-7.
Moretti P, Freeman K, Coodly L, Shore D. Evidence that a complex of SIR proteins interacts
with the silencer and telomere-binding protein RAP1. Genes Dev. 1994 Oct 1;8(19):2257-69.
Moretti P, Shore D. Multiple interactions in Sir protein recruitment by Rap1p at silencers and
telomeres in yeast. Mol Cell Biol. 2001 Dec;21(23):8082-94. Mumberg D, Müller R, Funk M. Yeast vectors for the controlled expression of heterologous
proteins in different genetic backgrounds. Gene. 1995 Apr 14;156(1):119-22. de Murcia, G., A. Huletsky, D. Lamarre, A. Gaudreau, J. Pouyet, M. Daune, and G. G. Poirier.
1986. Modulation of chromatin superstructure induced by poly(ADP-ribose) synthesis and degradation. J. Biol. Chem. 261:7011–7017.
Muthurajan UM, Park YJ, Edayathumangalam RS, Suto RK, Chakravarthy S, Dyer PN, Luger K.
Structure and dynamics of nucleosomal DNA. Biopolymers. 2003 Apr;68(4):547-56. Murphy DJ, Hardy S, Engel DA. Human SWI-SNF component BRG1 represses transcription of
the c-fos gene. Mol Cell Biol. 1999 Apr;19(4):2724-33. Muro-Pastor MI, Barrera FN, Reyes JC, Florencio FJ, Neira JL. The inactivating factor of
glutamine synthetase, IF7, is a "natively unfolded" protein. Protein Sci. 2003 Jul;12(7):1443-54.
152
Nakagawa T, Kajitani T, Togo S, Masuko N, Ohdan H, Hishikawa Y, Koji T, Matsuyama T, Ikura T, Muramatsu M, Ito T. Deubiquitylation of histone H2A activates transcriptional initiation via trans-histone cross-talk with H3K4 di- and trimethylation. Genes Dev. 2008 Jan 1;22(1):37-49.
Nakamura TM, Du LL, Redon C, Russell P. Histone H2A phosphorylation controls Crb2
recruitment at DNA breaks, maintains checkpoint arrest, and influences DNA repair in fission yeast. Mol Cell Biol. 2004 Jul;24(14):6215-30.
Nakayama J, Rice JC, Strahl BD, Allis CD, Grewal SI. Role of histone H3 lysine 9 methylation
in epigenetic control of heterochromatin assembly. Science. 2001 Apr 6;292(5514):110-3.
Nan X, Campoy FJ, Bird A. MeCP2 is a transcriptional repressor with abundant binding sites in
genomic chromatin. Cell. 1997 Feb 21;88(4):471-81. Nan X, Ng HH, Johnson CA, Laherty CD, Turner BM, Eisenman RN, Bird A. Transcriptional
repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex. Nature. 1998 May 28;393(6683):386-9.
Nanduri S, Carpick BW, Yang Y, Williams BR, Qin J. Structure of the double-stranded RNA-
binding domain of the protein kinase PKR reveals the molecular basis of its dsRNA-mediated activation. EMBO J. 1998 Sep 15;17(18):5458-65.
Nathan D, Ingvarsdottir K, Sterner DE, Bylebyl GR, Dokmanovic M, Dorsey JA, Whelan KA,
Krsmanovic M, Lane WS, Meluh PB, Johnson ES, Berger SL. Histone sumoylation is a negative regulator in Saccharomyces cerevisiae and shows dynamic interplay with positive-acting histone modifications. Genes Dev. 2006;20(8):966–976.
Nielsen AL, Oulad-Abdelghani M, Ortiz JA, Remboutsika E, Chambon P, Losson R.
Heterochromatin formation in mammalian cells: interaction between histones and HP1 proteins. Mol Cell. 2001 Apr;7(4):729-39.
Nelson CJ, Santos-Rosa H, Kouzarides T. Proline isomerization of histone H3 regulates lysine
methylation and gene expression. Cell. 2006 Sep 8;126(5):905-16. Nikitina T, Ghosh RP, Horowitz-Scherer RA, Hansen JC, Grigoryev SA, Woodcock CL.
MeCP2-chromatin interactions include the formation of chromatosome-like structures and are altered in mutations causing Rett syndrome. J Biol Chem. 2007 Sep 21;282(38):28237-45.
Ng HH, Zhang Y, Hendrich B, Johnson CA, Turner BM, Erdjument-Bromage H, Tempst P,
Reinberg D, Bird A. MBD2 is a transcriptional repressor belonging to the MeCP1 histone deacetylase complex. Nat Genet. 1999 Sep;23(1):58-61.
153
Ohki, I., Shimotake, N., Fujita, N., Jee, J., Ikegami, T., Nakao, M., and Shirakawa, M. (2001) Cell 105, 487–497.
Onishi M, Liou GG, Buchberger JR, Walz T, Moazed D. Role of the conserved Sir3-BAH
domain in nucleosome binding and silent chromatin assembly. Mol Cell. 2007 Dec 28;28(6):1015-28.
Orrico A, Lam C, Galli L, Dotti MT, Hayek G, Tong SF, Poon PM, Zappella M, Federico A,
Sorrentino V. MECP2 mutation in male patients with non-specific X-linked mental retardation. FEBS Lett. 2000 Sep 22;481(3):285-8.
Paetkau DW, Riese JA, MacMorran WS, Woods RA, Gietz RD.Interaction of the yeast RAD7
and SIR3 proteins: implications for DNA repair and chromatin structure.Genes Dev. 1994 Sep 1;8(17):2035-45.
Palladino F, Laroche T, Gilson E, Axelrod A, Pillus L, Gasser SM.SIR3 and SIR4 proteins are
required for the positioning and integrity of yeast telomeres.Cell. 1993 Nov 5;75(3):543-55.
Panyutin IG, Hsieh P. Formation of a single base mismatch impedes spontaneous DNA branch
migration. J Mol Biol. 1993 Mar 20;230(2):413-24. Paull TT, Cortez D, Bowers B, Elledge SJ, Gellert M. Direct DNA binding by Brca1. Proc Natl
Acad Sci U S A. 2001 May 22;98(11):6086-91. Park PU, Defossez PA, Guarente L.Effects of mutations in DNA repair genes on formation of
ribosomal DNA circles and life span in Saccharomyces cerevisiae. Mol Cell Biol. 1999 May;19(5):3848-56.
Pelka GJ, Watson CM, Radziewic T, Hayward M, Lahooti H, Christodoulou J, Tam PP. Mecp2
deficiency is associated with learning and cognitive deficits and altered gene activity in the hippocampal region of mice. Brain. 2006 Apr;129(Pt 4):887-98.
Peterson CL, Herskowitz I. Characterization of the yeast SWI1, SWI2, and SWI3 genes, which
encode a global activator of transcription. Cell. 1992 Feb 7;68(3):573-83. Phelan ML, Schnitzler GR, Kingston RE. Octamer transfer and creation of stably remodeled
nucleosomes by human SWI-SNF and its isolated ATPases. Mol Cell Biol. 2000 Sep;20(17):6380-9.
Pil PM, Lippard SJ. Specific binding of chromosomal protein HMG1 to DNA damaged by the
anticancer drug cisplatin. Science. 1992 Apr 10;256(5054):234-7. Pillus L, Rine J. Epigenetic inheritance of transcriptional states in S. cerevisiae. Cell. 1989 Nov
17;59(4):637-47.
154
Probst AV, Dunleavy E, Almouzni G. Epigenetic inheritance during the cell cycle. Nat Rev Mol Cell Biol. 2009 Mar;10(3):192-206.
Prokhortchouk A, Hendrich B, Jørgensen H, Ruzov A, Wilm M, Georgiev G, Bird A,
Prokhortchouk E. The p120 catenin partner Kaiso is a DNA methylation-dependent transcriptional repressor. Genes Dev. 2001 Jul 1;15(13):1613-8.
Prokhortchouk A, Sansom O, Selfridge J, Caballero IM, Salozhin S, Aithozhina D, Cerchietti L,
Meng FG, Augenlicht LH, Mariadason JM, Hendrich B, Melnick A, Prokhortchouk E, Clarke A, Bird A. Kaiso-deficient mice show resistance to intestinal cancer. Mol Cell Biol. 2006 Jan;26(1):199-208.
Pruss D, Bushman FD, Wolffe AP. Human immunodeficiency virus integrase directs integration.
to sites of severe DNA distortion within the nucleosome core. Proc Natl Acad Sci U S A. 1994 Jun 21;91(13):5913-7.
Pryciak PM, Varmus HE. Nucleosomes, DNA-binding proteins, and DNA sequence modulate
retroviral integration target site selection. Cell. 1992 May 29;69(5):769-80. Pullen RA, Jenkins JA, Tickle IJ, Wood SP, Blundell TL. The relation of polypeptide hormone
structure and flexibility to receptor binding: the relevance of X-ray studies on insulins, glucagon and human placental lactogen. Mol Cell Biochem. 1975 Jul 31;8(1):5-20.
Racki LR, Narlikar GJ. ATP-dependent chromatin remodeling enzymes: two heads are not
better, just different. Curr Opin Genet Dev. 2008 Apr;18(2):137-44. Ray A, Hector RE, Roy N, Song JH, Berkner KL, Runge KW.Sir3p phosphorylation by the Slt2p
pathway effects redistribution of silencing function and shortened lifespan. Nat Genet. 2003 Apr;33(4):522-6. Epub 2003 Mar 17. Erratum in: Nat Genet. 2003 May;34(1):113.
Reeves R, Nissen MS. The A.T-DNA-binding domain of mammalian high mobility group I
chromosomal proteins. A novel peptide motif for recognizing DNA structure. J Biol Chem. 1990 May 25;265(15):8573-82.
Reeves R, Wolffe AP. Substrate structure influences binding of the non-histone protein HMG-
I(Y) to free nucleosomal DNA. Biochemistry. 1996 Apr 16;35(15):5063-74. Renauld H, Aparicio OM, Zierath PD, Billington BL, Chhablani SK, Gottschling DE.Silent
domains are assembled continuously from the telomere and are defined by promoter distance and strength, and by SIR3 dosage. Genes Dev. 1993 Jul;7(7A):1133-45.
Rice JC, Briggs SD, Ueberheide B, Barber CM, Shabanowitz J, Hunt DF, Shinkai Y, Allis CD.
Histone methyltransferases direct different degrees of methylation to define distinct chromatin domains. Mol Cell. 2003 Dec;12(6):1591-8.
155
Rine J, Herskowitz I. Four genes responsible for a position effect on expression from HML and HMR in Saccharomyces cerevisiae. Genetics. 1987 May;116(1):9-22.
Robert CH, Gill SJ, and Wyman J. Quantitative analysis of linkage in macromolecules when one
ligand is present in limited total quantity. Biochemistry. 1988;27: 6829-6835. Rogakou EP, Pilch DR, Orr AH, Ivanova VS, Bonner WM. DNA double-stranded breaks induce
histone H2AX phosphorylation on serine 139. J. Biol. Chem. 1998;273:5858–5868. Romero P, Obradovic Z, Li X, Garner EC, Brown CJ, Dunker AK. Sequence complexity of
disordered protein. Proteins. 2001 Jan 1;42(1):38-48. Rusche LN, Kirchmaier AL, Rine J. The establishment, inheritance, and function of silenced
chromatin in Saccharomyces cerevisiae.Annu Rev Biochem. 2003;72:481-516. Ruzov A, Dunican DS, Prokhortchouk A, Pennings S, Stancheva I, Prokhortchouk E, Meehan
RR. Kaiso is a genome-wide repressor of transcription that is essential for amphibian development. Development. 2004 Dec;131(24):6185-94.
Sandell LL, Zakian VA. Telomeric position effect in yeast. Trends Cell Biol. 1992 Jan;2(1):10-4. Sansom OJ, Berger J, Bishop SM, Hendrich B, Bird A, Clarke AR. Deficiency of Mbd2
suppresses intestinal tumorigenesis. Nat Genet. 2003 Jun;34(2):145-7. Sarraf SA, Stancheva I. Methyl-CpG binding protein MBD1 couples histone H3 methylation at
lysine 9 by SETDB1 to DNA replication and chromatin assembly. Mol Cell. 2004 Aug 27;15(4):595-605.
Schulz FE. Nucleotide binding proteins. In Molecular Mechanism of Biological Recognition,
Balaban M. Elsevier/North-Holland Biomedical Press: New York, 1979: 79-94. Scully R, Chen J, Ochs RL, Keegan K, Hoekstra M, Feunteun J, Livingston DM. Dynamic
changes of BRCA1 subnuclear location and phosphorylation state are initiated by DNA damage. Cell. 1997 Aug 8;90(3):425-35.
Shi Y, Lan F, Matson C, Mulligan P, Whetstine JR, Cole PA, Casero RA, Shi Y. Histone
demethylation mediated by the nuclear amine oxidase homolog LSD1. Cell. 2004;119:941–953.
Shibayama A, Cook EH Jr, Feng J, Glanzmann C, Yan J, Craddock N, Jones IR, Goldman D,
Heston LL, Sommer SS. MECP2 structural and 3'-UTR variants in schizophrenia, autism and other psychiatric diseases: a possible association with autism. Am J Med Genet B Neuropsychiatr Genet. 2004 Jul 1;128B(1):50-3.
Shore D, Squire M, Nasmyth KA. Characterization of two genes required for the position-effect
control of yeast mating-type genes. EMBO J. 1984 Dec 1;3(12):2817-23.
156
Shore D, Nasmyth K.Purification and cloning of a DNA binding protein from yeast that binds to
both silencer and activator elements. Cell. 1987 Dec 4;51(5):721-32. Simon J, Chiang A, Bender W. Ten different Polycomb group genes are required for spatial
control of the abdA and AbdB homeotic products. Development. 1992 Feb;114(2):493-505.
Smith JS, Brachmann CB, Celic I, Kenna MA, Muhammad S, Starai VJ, Avalos JL, Escalante-
Semerena JC, Grubmeyer C, Wolberger C, Boeke JD.A phylogenetically conserved NAD+-dependent protein deacetylase activity in the Sir2 protein family.Proc Natl Acad Sci U S A. 2000 Jun 6;97(12):6658-63.
Springhetti EM, Istomina NE, Whisstock JC, Nikitina T, Woodcock CL, Grigoryev SA. Role of
the M-loop and reactive center loop domains in the folding and bridging of nucleosome arrays by MENT. J Biol Chem. 2003 Oct 31;278(44):43384-93.
Spolar RS, Record MT Jr. Coupling of local folding to site-specific binding of proteins to DNA.
Science. 1994 Feb 11;263(5148):777-84. Stone EM, Pillus L. Silent chromatin in yeast: an orchestrated medley featuring Sir3p. Bioessays.
1998 Jan;20(1):30-40. Strahl BD, Allis CD. The language of covalent histone modifications. Nature. 2000 Jan
6;403(6765):41-5. Strahl-Bolsinger S, Hecht A, Luo K, Grunstein M. SIR2 and SIR4 interactions differ in core and
extended telomeric heterochromatin in yeast. Genes Dev. 1997 Jan 1;11(1):83-93. Struhl G. A gene product required for correct initiation of segmental determination in
Drosophila. Nature. 1981 Sep 3;293(5827):36-41. Steinmetz EJ, Warren CL, Kuehner JN, Panbehi B, Ansari AZ, Brow DA. Genome-wide
distribution of yeast RNA polymerase II and its control by Sen1 helicase. Mol Cell. 2006 Dec 8;24(5):735-46.
Suka N, Suka Y, Carmen AA, Wu J, Grunstein M. Highly specific antibodies determine histone
acetylation site usage in yeast heterochromatin and euchromatin. Mol Cell. 2001 Aug;8(2):473-9.
Suzuki M, Yamada T, Kihara-Negishi F, Sakurai T, Oikawa T. Direct association between PU.1
and MeCP2 that recruits mSin3A-HDAC complex for PU.1-mediated transcriptional repression. Oncogene. 2003 Nov 27;22(54):8688-98.
Tanny JC, Dowd GJ, Huang J, Hilz H, Moazed D. An enzymatic activity in the yeast Sir2 protein
that is essential for gene silencing. Cell. 1999 Dec 23;99(7):735-45.
157
Tao J, Hu K, Chang Q, Wu H, Sherman NE, Martinowich K, Klose RJ, Schanen C, Jaenisch R,
Wang W, Sun YE. Phosphorylation of MeCP2 at Serine 80 regulates its chromatin association and neurological function. Proc Natl Acad Sci U S A. 2009 Mar 24;106(12):4882-7.
Teif VB. General transfer matrix formalism to calculate DNA-protein-drug binding in gene
regulation: application to OR operator of phage lambda. Nucleic Acids Res. 2007;35(11):e80.
Thomas JO, Rees C, Finch JT. Cooperative binding of the globular domains of histones H1 and
H5 to DNA. Nucleic Acids Res. 1992 Jan 25;20(2):187-94. Topalova D, Ugrinova I, Pashev IG, Pasheva EA. HMGB1 protein inhibits DNA replication in
vitro: A role of the acetylation and the acidic tail. Int J Biochem Cell Biol. 2008;40(8):1536-42.
Trifonov EN. DNA in profile. Trends Biochem Sci. 1991 Dec;16(12):467-70. Trouche D, Le Chalony C, Muchardt C, Yaniv M, Kouzarides T. RB and hbrm cooperate to
repress the activation functions of E2F1. Proc Natl Acad Sci U S A. 1997 Oct 14;94(21):11268-73.
Tse C, Georgieva EI, Ruiz-García AB, Sendra R, Hansen JC. Gcn5p, a transcription-related histone acetyltransferase, acetylates nucleosomes and folded nucleosomal arrays in the absence of other protein subunits. J Biol Chem. 1998 Dec 4;273(49):32388-92.
Ueda T, Chou H, Kawase T, Shirakawa H, Yoshida M. Acidic C-tail of HMGB1 is required for its target binding to nucleosome linker DNA and transcription stimulation. Biochemistry. 2004 Aug 3;43(30):9901-8.
Uversky VN, Oldfield CJ, Dunker AK. Showing your ID: intrinsic disorder as an ID for
recognition, regulation and cell signaling. J Mol Recognit. 2005 Sep-Oct;18(5):343-84. Vakoc CR, Mandat SA, Olenchock BA, Blobel GA. Histone H3 lysine 9 methylation and
HP1gamma are associated with transcription elongation through mammalian chromatin. Mol Cell. 2005 Aug 5;19(3):381-91.
Van den Veyver IB, Zoghbi HY. Mutations in the gene encoding methyl-CpG-binding protein 2
cause Rett syndrome. Brain Dev. 2001 Dec;23 Suppl 1:S147-51. Review. Van Esch H, Bauters M, Ignatius J, Jansen M, Raynaud M, Hollanders K, Lugtenberg D,
Bienvenu T, Jensen LR, Gecz J, Moraine C, Marynen P, Fryns JP, Froyen G. Duplication of the MECP2 region is a frequent cause of severe mental retardation and progressive neurological symptoms in males. Am J Hum Genet. 2005 Sep;77(3):442-53.
158
Varga-Weisz P, Zlatanova J, Leuba SH, Schroth GP, van Holde K. Binding of histones H1 and H5 and their globular domains to four-way junction DNA. Proc Natl Acad Sci U S A. 1994 Apr 26;91(9):3525-9.
Verdel A, Jia S, Gerber S, Sugiyama T, Gygi S, Grewal SI, Moazed D. RNAi-mediated targeting
of heterochromatin by the RITS complex. Science. 2004 Jan 30;303(5658):672-6. Verschure PJ, van der Kraan I, de Leeuw W, van der Vlag J, Carpenter AE, Belmont AS, van
Driel R. In vivo HP1 targeting causes large-scale chromatin condensation and enhanced histone lysine methylation. Mol Cell Biol. 2005 Jun;25(11):4552-64.
Virta-Pearlman V, Morris DK, Lundblad V. Est1 has the properties of a single-stranded telomere
end-binding protein. Genes Dev. 1996 Dec 15;10(24):3094-104. Volpe TA, Kidner C, Hall IM, Teng G, Grewal SI, Martienssen RA. Regulation of
heterochromatic silencing and histone H3 lysine-9 methylation by RNAi. Science. 2002 Sep 13;297(5588):1833-7.
Wade PA, Gegonne A, Jones PL, Ballestar E, Aubry F, Wolffe AP. Mi-2 complex couples DNA
methylation to chromatin remodelling and histone deacetylation. Nat Genet. 1999 Sep;23(1):62-6.
Wakefield RI, Smith BO, Nan X, Free A, Soteriou A, Uhrin D, Bird AP, Barlow PN. The
solution structure of the domain from MeCP2 that binds to methylated DNA. J Mol Biol. 1999 Sep 3;291(5):1055-65.
Wang H, Wang L, Erdjument-Bromage H, Vidal M, Tempst P, Jones RS, Zhang Y. Role of
histone H2A ubiquitination in Polycomb silencing. Nature. 2004 Oct 14;431(7010):873-8.
Watson P, Black G, Ramsden S, Barrow M, Super M, Kerr B, Clayton-Smith J. Angelman
syndrome phenotype associated with mutations in MECP2, a gene encoding a methyl CpG binding protein. J Med Genet. 2001 Apr;38(4):224-8.
Watson M, Stott K, Thomas JO. Mapping intramolecular interactions between domains in
HMGB1 using a tail-truncation approach. J Mol Biol. 2007 Dec 14;374(5):1286-97. Weintraub H, Groudine M. Chromosomal subunits in active genes have an altered conformation.
Science. 1976 Sep 3;193(4256):848-56. Weiss MA, Ellenberger T, Wobbe CR, Lee JP, Harrison SC, Struhl K. Folding transition in the
DNA-binding domain of GCN4 on specific binding to DNA. Nature. 1990 Oct 11;347(6293):575-8.
159
Wen L, Huang JK, Johnson BH, Reeck GR. A human placental cDNA clone that encodes nonhistone chromosomal protein HMG-1. Nucleic Acids Res. 1989 Feb 11;17(3):1197-214.
Wischnewski F, Friese O, Pantel K, Schwarzenbach H. Methyl-CpG binding domain proteins
and their involvement in the regulation of the MAGE-A1, MAGE-A2, MAGE-A3, and MAGE-A12 gene promoters. Mol Cancer Res. 2007 Jul;5(7):749-59.
Wolffe AP Chromtin: Structure and Function. 1998 Wong E, Yang K, Kuraguchi M, Werling U, Avdievich E, Fan K, Fazzari M, Jin B, Brown AM,
Lipkin M, Edelmann W. Mbd4 inactivation increases Cright-arrowT transition mutations and promotes gastrointestinal tumor formation. Proc Natl Acad Sci U S A. 2002 Nov 12;99(23):14937-42.
Wutz A., Rasmussen, T.P., and Jaenisch, R. 2002. Chromosomal silencing and localization are
mediated by different domains of Xist RNA. Nat. Genet. 30: 167–174. Wyman J, and Gill SJ. Binding and Linkage, University Science Books. 1990. Hill Valley, CA. Xu EY, Bi X, Holland MJ, Gottschling DE, Broach JR. Mutations in the nucleosome core
enhance transcriptional silencing. Mol Cell Biol. 2005 Mar;25(5):1846-59. Yang X, Zaurin R, Beato M, Peterson CL. Swi3p controls SWI/SNF assembly and ATP-
dependent H2A-H2B displacement. Nat Struct Mol Biol. 2007 Jun;14(6):540-7. Yasui DH, Peddada S, Bieda MC, Vallero RO, Hogart A, Nagarajan RP, Thatcher KN, Farnham
PJ, Lasalle JM. Integrated epigenomic analyses of neuronal MeCP2 reveal a role for long-range interaction with active genes. Proc Natl Acad Sci U S A. 2007 Dec 4;104(49):19416-21.
Yoon HG, Chan DW, Reynolds AB, Qin J, Wong J. N-CoR mediates DNA methylation-
dependent repression through a methyl CpG binding protein Kaiso. Mol Cell. 2003 Sep;12(3):723-34.
Young JI, Hong EP, Castle JC, Crespo-Barreto J, Bowman AB, Rose MF, Kang D, Richman R,
Johnson JM, Berget S, Zoghbi HY. Regulation of RNA splicing by the methylation-dependent transcriptional repressor methyl-CpG binding protein 2. Proc Natl Acad Sci U S A. 2005 Dec 6;102(49):17551-8.
Yusufzai TM, Wolffe AP. Functional consequences of Rett syndrome mutations on human
MeCP2. Nucleic Acids Res. 2000 Nov 1;28(21):4172-9. Zhang H, Somasundaram K, Peng Y, Tian H, Zhang H, Bi D, Weber BL, El-Deiry WS. BRCA1
physically associates with p53 and stimulates its transcriptional activity. Oncogene. 1998 Apr 2;16(13):1713-21.
160
Zhang Y, LeRoy G, Seelig HP, Lane WS, Reinberg D. The dermatomyositis-specific autoantigen
Mi2 is a component of a complex containing histone deacetylase and nucleosome remodeling activities. Cell. 1998 Oct 16;95(2):279-89.
Zhao X, Ueba T, Christie BR, Barkho B, McConnell MJ, Nakashima K, Lein ES, Eadie BD,
Willhoite AR, Muotri AR, Summers RG, Chun J, Lee KF, Gage FH. Mice lacking methyl-CpG binding protein 1 have deficits in adult neurogenesis and hippocampal function. Proc Natl Acad Sci U S A. 2003 May 27;100(11):6777-82.
Zhao Y, Lang G, Ito S, Bonnet J, Metzger E, Sawatsubashi S, Suzuki E, Le Guezennec X,
Stunnenberg HG, Krasnov A, Georgieva SG, Schüle R, Takeyama K, Kato S, Tora L, Devys D. A TFTC/STAGA module mediates histone H2A and H2B deubiquitination, coactivates nuclear receptors, and counteracts heterochromatin silencing. Mol Cell. 2008 Jan 18;29(1):92-101.
Zheng C, Lu X, Hansen JC, Hayes JJ. Salt-dependent intra- and internucleosomal interactions of
the H3 tail domain in a model oligonucleosomal array. J Biol Chem. 2005 Sep 30;280(39):33552-7.
Zhou W, Wang X, Rosenfeld MG. Histone H2A ubiquitination in transcriptional regulation and
DNA damage repair. Int J Biochem Cell Biol. 2009 Jan;41(1):12-5. Zhong Q, Chen CF, Li S, Chen Y, Wang CC, Xiao J, Chen PL, Sharp ZD, Lee WH. Association
of BRCA1 with the hRad50-hMre11-p95 complex and the DNA damage response. Science. 1999 Jul 30;285(5428):747-50.
Zhou Z, Hong EJ, Cohen S, Zhao WN, Ho HY, Schmidt L, Chen WG, Lin Y, Savner E, Griffith
EC, Hu L, Steen JA, Weitz CJ, Greenberg ME. Brain-specific phosphorylation of MeCP2 regulates activity-dependent Bdnf transcription, dendritic growth, and spine maturation. Neuron. 2006 Oct 19;52(2):255-69.
Zhou W, Zhu P, Wang J, Pascual G, Ohgi KA, Lozach J, Glass CK, Rosenfeld MG. Histone
H2A monoubiquitination represses transcription by inhibiting RNA polymerase II transcriptional elongation. Mol Cell. 2008 Jan 18;29(1):69-80.
Zhu N, Hansen U. HMGN1 modulates estrogen-mediated transcriptional activation through
interactions with specific DNA-binding transcription factors. Mol Cell Biol. 2007 Dec;27(24):8859-73.