Inhibition of IFN-y Promoter Function by Site-Specific Methylation
by
MASsAcHUJSEr-IS1TrrEI
OF TECHNOLOGY
Brendan T. Jones
FEB
8
1 2006
B.S. Biochemistry, with High Honors
Tufts University, 1999
LIBRARIE
Submitted to the Department of Biology in
Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy
'1
at the
C
IVES 8)
Massachusetts Institute of Technology
gro bvcAail, ooG 3
January 2006
© 2006 Massachusetts Institute of Technology
All rights reserved
Signature of Author:
,,,/
-
IA,
I
Brendan T. Jones
Department of Biology
January 4, 2006
Certified by:
Jianzhu Chen
Professor of Biology
January 4, 2006
Accepted by:
t " -It-
Stephen P. Bell
Professor of Biology
January 4, 2006
1
Inhibition of IFN-y Promoter Function by Site-Specific Methylation
by
Brendan T. Jones
Submitted to the Department of Biology on December 13, 2005 in Partial Fulfillment of
the Requirements for the Degree of Doctor of Philosophy
ABSTRACT
When they become activated, na'ive helper T cells are able to polarize into either
THI cells or TH2 cells. Development of naive CD4 + T cells into TH1 cells is characterized
by the expression of IFN-y and the silencing of IL-4, while development into TH2 cells is
characterized by expression of IL-4 and silencing of IFN-y. NaYve helper T cells are
hypomethylated at the IFN-y proximal promoter and hypermethylated at the contiguous
transcribed region. During THI polarization, the promoter remains hypomethylated,
while the transcribed region becomes demethylated. During TH2 polarization, the
promoter undergoes progressive de novo methylation while the transcribed region
remains hypermethylated. Notably, TH2 de novo methylation occurs at different rates at
different CpG positions, with methylation occurring fastest at the CpG located at the -53
position relative to the transcription start site. Methylation at this position inhibits c-Jun,
ATF2 and CREB binding in vitro. Consistently, the same factors bind to the
unmethylated
promoter in a TH1 cell line, but not the methylated promoter in a TH2 cell
line. Furthermore, methylation of the proximal promoter at the -53 position alone is
sufficient to inhibit promoter activity in transient transfection assays. Thus, the rapid
methylation of the -53 CpG at the onset of TH2 polarization helps to prevent IFN-y
transcription by directly inhibiting transcription factor binding prior to the extensive
mnethylation of the IFN-y promoter.
There are three known mammalian methyltransferase genes: dnmtl, dnmt3a, and
dcnmt3b. Dnmt3b is not required for the methylation changes that occur at the IFN-y
locus during helper T cell polarization. De novo methylation during TH2 polarization is
reduced in dnmt3a deficient T cells. Furthermore, helper T cells deficient in the dnmt3a
alternative transcript, dnmt3a2, undergo de novo methylation at the IFN-y promoter
during TH 1 polarization, and IFN-y expression is inhibited in these T cells. Collectively,
this suggests that dnmt3a is required for efficient de novo methylation of the IFN-y
promoter during TH2 polarization, and that dnmt3a2 suppresses IFN-y methylation during
TH1 polarization.
Thesis Supervisor: Jianzhu Chen
Title: Professor of Biology
2
Acknowledgements
Many people were involved in the generation of this thesis, both directly and
indirectly. First and foremost, I would like to thank Jianzhu Chen, who has been an
excellent mentor and a good friend. Without his guidance and patience I never would
have succeeded here at MIT. He taught me how to think like a scientist. My thanks also
go to my thesis committee for their advice and support.
Many members of the Chen Lab have given me both technical and moral support
over the years. Qing Ge and Hui Hu taught me everything I know about mouse work. As
busy as they were, they always found time to help me find that last pesky lymph node.
Without Charles Whitehurst and Ching-Hung Shen, the bisulfite conversion and gel shift
assays would have been impossible. Many thanks also to Mimi Fragoso and Jim
Harriman for ordering my reagents and looking after my mice.
The people who have shared my bay over the years, Brian Haines, Ailin Bai, and
Lily Trajman all get special credit for putting up with both my music and my antics.
With less tolerant neighbors I would have surly ended up in cement shoes at the bottom
of the River Charles. An additional thank you to Lily for proofreading this thesis.
Without her grammatical expertise, the document you hold before you would have been
an unreadable jumble of verb disagreement and misplaced commas.
Finally need to thank my family: Courtney, for putting up with my long hours
and low pay, and for providing the loving support that kept me sane, my mother for
instilling in me the love of learning that drove me into science, Donovan for giving me a
reason to get home early, and Pauline for watching Donovan when I was unable to.
3
Table of Contents
Title Page
1
Abstract
2
Acknowledgements
3
Table of Contents
4
Chapter 1: Introduction
5
T cell Development and Activation
7
Helper T cell Polarization
14
Regulation of IFN-y Transcription
28
Epigenetic Regulation of Helper T cell Polarization
34
References
43
Chapter 2: Site-Specific Methylation Suppresses IFN-y Promoter
Activity by Directly Inhibiting Factor Binding
61
Summary
62
Introduction
63
Results
68
Discussion
107
Materials and Methods
115
References
122
Chapter 3: The Roles of Specific Methyltransferases in the
Methylation Changes at the IFN-y Locus during TH2Polarization
127
Summary
128
Introduction
129
Results
133
Discussion
155
Materials and Methods
161
References
163
Chapter 3: Further Discussion and Future Direction
166
4
Chapter 1:
Introduction
5
In many ways, the mammalian body is an ideal environment for pathogenic
infection.
Within mammals, microorganisms find reliable water, abundant energy,
plentiful nutrients, and a constant temperature. While there are many microorganisms
that are either beneficial or neutral to the host's wellbeing, there are countless more that
are harmful or even deadly. Even those viewed as harmless, like the natural flora of the
human intestine, can be lethal if allowed to grow unchecked.
For this reason it is
imperative that mammals have a robust and effective immune system.
While the clearance of invading pathogens is necessary for host survival, a poorly
regulated immune response can do more harm than good. An immune response that fails
to differentiate
between foreign microorganisms
and its own cells can lead to
autoimmune diseases like rheumatoid arthritis (in which the immune system mounts a
response against the synovium covering joints) (Goronzy and Weyand, 2005), multiple
sclerosis (in which immune cells attack the myelin sheath surrounding nerve fibers)
(Hafler et al., 2005), or type 1 diabetes mellitus (in which an immune response destroys
beta cells in the pancreatic islets of Langerhans.) (Kelly et al., 2003) The immune
system's failure to differentiate between pathogens and harmless environmental antigens,
like plant pollen or animal dander, often results in allergic responses or asthma. Even
when an immune response is properly directed against a harmful pathogen, an overly
aggressive and uncontrolled response can lead to extensive tissue damage and even death.
It is possible that this was the case in many of the influenza fatalities among young adults
during the 1918 pandemic (Luk et al., 2001).
For the above reasons, the immune system must maintain a delicate balance
whereby efficacy is maintained, but a harmful over-response is avoided. For this to
happen, foreign cells must be distinguished from the host's own cells, and harmful
pathogens must be distinguished from harmless environmental antigens. The response
must be robust enough to clear infections, but must also be controlled and limited so as to
minimize harm to the host. If these balances are not maintained it is likely that the host
will die, either by being overrun by the pathogenic infection or by succumbing to its own
immune response.
This strong selective pressure has lead to the evolution of a
remarkably effective mammalian immune system made up of millions of diverse,
6
interactive cells capable of a coordinated and specific response against pathogenic
infection.
T CELL DEVELOPMENT AND ACTIVATION
T cell Development
The mammalian immune system can be divided into two interacting components,
the innate immune system and the adaptive immune system. The innate immune system
is able to mount a rapid, relatively non-specific response against pathogens that are easily
recognized as foreign.
In order to do this, the innate immune system relies on the
presence of pathogen associated molecular patterns (PAMPs), which are molecules that
are frequently found in common pathogens, but are rarely, if ever, found in mammalian
cells. Some examples of PAMPs include the flagellin from bacterial flagella (Hayashi et
al., 2001), lipopolysaccharide (LPS) from the cell wall of gram-negative bacteria
(Yoshimura et al., 1999), and viral double-stranded RNA (Alexopoulou et al., 2001).
While the innate immune system can be an effective first line of defense against many
types of pathogens, it relies on a relatively small number of pathogen specific molecules
to distinguish foreign from self and a launches a non-specific response.
In addition to providing a nonspecific response to PAMPs, the innate immune
system also plays a role in the priming of the adaptive immune system, which is capable
of more specific and directed responses. The two major cell types that make up the
adaptive immune system are T and B lymphocytes. The distinguishing surface protein of
B3cells is the B cell receptor (BCR), while the distinguishing surface protein of T cells in
the T cell receptor (TCR.) Both the BCR and the TCR are responsible for the recognition
of foreign antigens. In order to recognize a wide variety of antigens, the antigen
recognition regions of the TCR and BCR are generated through the process of VDJ
recombination.
VDJ recombination is a process, mediated by the RAG recombinase
enzymes, in which the receptor loci are rearranged at the DNA level, resulting in each
cell producing a unique receptor (Alt et al., 1992; Fugmann, 2001).
This allows the T
cell and B cell compartments to recognize millions of different antigens. While B cells
are able to recognize a wide range of soluble antigens through their BCR (Pleiman et al.,
7
1994), T cells, through their TCR, recognize short, processed peptide antigens presented
on major histocompatibility
complex (MHC) proteins on the surface of antigen
presenting cells (APCs) (Garcia et al., 1999). An antigen-specific immune response is
initiated when T cells and B cells undergo clonal expansion after encountering antigen
that is recognized by their TCR or BCR respectively.
Both T and B cells are derived from a common lymphoid precursor in the bone
marrow (Hirose et al., 2002) (Figure 1). While B cells continue to mature in the bone
marrow, progenitor T cells leave the bone marrow early in development and migrate to
the thymus (Shortman and Wu, 1996). During the early stages of thymic development,
the T cell precursors express neither of the T cell coreceptors, CD4 and CD8 (Wu et al.,
1991). For this reason, these early thymocytes are referred to as CD4, CD8 double
negative (DN) thymocytes.
During the double negative stage of thymic development thymocytes begin to
express the RAG recombinase enzymes and to rearrange the m chain of their TCR
(Godfrey et al., 1994). Once a productive rearrangement has occurred on one of the
alleles, the cell is able produce TCR3 protein. This protein is able to dimerize with a preTCRct protein, and is expressed on the thymocyte surface with the CD3 proteins to form
the pre-TCR complex (Saint-Ruf et al., 1994). Surface expression of this complex sends
a signal to the DN thymocyte that results in a cessation of TCR3 rearrangement, a burst
of proliferation, and the upregulation of CD4 and CD8 coreceptors (von Boehmer, 2005).
At this point the cell becomes a CD4, CD8 double positive (DP) thymocyte.
DP thymocytes also express the RAG recombinase enzymes and rearrange the
TCRco locus (Sleckman et al., 1998). Once productive rearrangement has occurred, the
TCR1 chain and TCRca chain are able to form a heterodimer and, along with the CD3
proteins, are expressed on the cell surface as a TCR complex (Sleckman et al., 1998). It
is during the DP stage of thymic development that the thymocytes undergo positive and
negative selection. In order to survive, the TCR on DP thymocytes must be able to bind
to cortical epithelial cell-expressed MHC Class I or MHC Class II complexed with selfpeptide (von Boehmer et al., 2003). If the thymocyte's TCR is unable to bind, the cells
die via apoptosis due to a lack of survival signals. On the other hand, if the TCR on the
DP thymocyte binds too tightly to a dendritic cell (DC) or macrophage expressed self-
8
Bone Marrow
Common
.ymphoid
Natural
Killer Cell
Cell
Memory
CD8 +
T Cell
9
Effector
CD8+
T Cell
JaIve
IaYve CD8
h elper
T Cell
I
+
Cell
Thymus
CD4
(( )Negative
Positive
rhymocyte
Thymocyte
Thyme
~~~,,o,,
N.
CD8
Positive
Thymocyte
- ..
L)
ouble
Positive
Ohymocyte
Figure 1. The T cell development pathway.
9
peptide presenting MHC, the thymocyte also undergoes apoptosis (Palmer, 2003). This
process ensures T cell self-tolerance, and is an important step in the prevention of
autoimmune disease.
DP thymocytes that survive positive and negative selection shut down expression
of either CD8 or CD4, becoming either CD4 or CD8 single positive thymocytes
respectively (Basson and Zamoyska, 2000).
CD8+ thymocytes recognize peptides
presented on MHC Class I and become cytotoxic T cells. These cells are important for
cellular immunity and are responsible for killing virally infected or cancerous cells.
CD4+ thymocytes recognize peptides presented on MHC Class II and become helper T
cells. Helper T cells enhance the immune response largely through cytokine secretion.
Once committed to either a CD4 or a CD8 lineage, the cells leave the thymus, becoming
mature, peripheral T cells (Figure 1).
Antigen Presentation and T cell Activation
Unlike the BCR, the TCR are unable to recognize soluble protein, and only react
to processed peptide complexed with MHC (Garcia et al., 1999). CD4+ T cells interact
with peptide presented on class II MHC, while CD8+ T cells interact with peptide
presented on class I MHC. For this reason, antigen processing and presentation, which
primarily done by innate immune cells, are important steps in the induction of an
adaptive immune response.
Class I MHC are expressed on the surface of all nucleated cells, and primarily
present endogenous peptides (Gromme and Neefjes, 2002).
Peptides destined for
presentation on Class I MHC are generated through the degradation of cytosolic proteins
by the proteasome complex (Kloetzel, 2004). Peptides that are between 8 and 10 amino
acids in length are preferentially transported from the cytosol to the rough endoplasmic
reticulum (RER) by TAP in an ATP dependant process. Once in the RER, these peptides
are able to complex with, and in so doing stabilize, class I MHC. This leads to the class I
MHC/peptide complex being expressed on the cell surface via the Golgi complex
(Gromme and Neefjes, 2002; Kloetzel, 2004). Because class I MHC express endogenous
peptides, this is the mechanism by which viral antigens are presented by infected cells
10
and tumor antigens are presented by transformed cells, rendering the cells targets for
killing by antigen-specific cytotoxic T cells.
Class II MHC are expressed on the surface of professional APCs, such as
dendritic cells, macrophages and B cells, and primarily complex with exogenous peptides
(Watts, 2004).
Antigens that have been internalized by either phagocytosis or
endocytosis are degraded into peptides within the endocytic processing pathway. Like
the class I MHC, class II MHC is synthesized within the RER, but the peptide-binding
region of class II MHC is blocked by a protein complex called the invariant chain,
preventing it from complexing with endogenous antigens (Villadangos and Ploegh,
2000). The class II MHC/invariant chain complex exits the RER via the Golgi complex
and into the endocytic pathway.
In the endocytic pathway, the invariant chain is
degraded, leaving only a small fragment bound to the class II MHC called CLIP. CLIP is
then displaced by exogenous peptide and the class I MHC/peptide complex is presented
on the cell surface, where it is able to interact with antigen-specific helper T cells
(Villadangos and Ploegh, 2000; Watts, 2004).
When CD4+ and CD8+ T cells leave the thymus they express a naive phenotype
(Figure 1). Nafve T cells are small, resting, and have very little metabolic activity. In
order to take part in an immune response, naive T cells must become activated and
differentiate into effector T cells.
In addition to mediating the immune response,
activated T cells rapidly proliferate, which results in clonal expansion.
Following
pathogen clearance, most of the activated T cells die by apoptosis, but a small fraction of
them differentiate into memory T cells, which are small, slow to divide and long lived.
Memory T cells can persist in a host organism for many years, and are able to mediate a
rapid and robust secondary immune response in the event of antigen reencounter (Seder
and Ahmed, 2003).
The activation of naive T cells requires at least two signals, one of which is
provided by the interaction of the TCR complex with peptide/MHC. This interaction
results in a cascade of signals, some of which are diagrammed in Figure 2. The TCR
itself lacks any signal transduction activity, and thus relies on the associated CD3y, 6, a,
proteins for signaling (Cantrell, 1996; Wange and Samelson, 1996).
The strong
interaction between peptide/MHC and TCR results in a clustering of TCR and either CD4
11
TCR/CD3
CD4/CD8
,
',FIJI
(CNF-K
MAP kinase P athway
f
1
Nucleus
Figure 2. Some of the molecular signals involved in
the TCR signal cascade.
12
or CD8 co-receptors at the immune synapse (Kupfer and Singer, 1989). The cytoplasmic
tail of the co-receptor molecules is associated with the protein tyrosine kinase ck, which,
following immune synapse formation, phosphorylates the immunoreceptor tyrosine-based
motifs (ITAMs) on the CD3 cytoplasmic tails (Straus and Weiss, 1992). Phosphorylation
of the ITAMs on CD3
creates a docking site for another protein tyrosine kinase, 4-
chain-associated protein 70 (ZAP-70), which is itself phosphorylated and activated by lck
(Chan et al., 1995; Kong et al., 1996; Wange et al., 1995).
Activated ZAP-70
phosphorylates a number of proteins, including phospholipase Cyl (PLCy1), SH2domain-containing leukocyte protein of 76 kDa (SLP-76) and linker for activation of T
cells (LAT) (Samelson, 2002; Zhang et al., 1998b). LAT is a transmembrane adaptor
protein and serves as a scaffold to mediate the recruitment of molecules involved in
several downstream signaling pathways, including PLCy1 and SLP-76 (Samelson, 2002;
Zhang et al., 1998b). Phosphorylated PLCy1 mediates the cleavage of the membrane
phospholipids phosphatidylinositol-(4,5)-bisphosphate
into diacylglycerol (DAG) and
inositol phosphate (IP3 ) (Carpenter and Ji, 1999). IP3 induces the release of intracellular
calcium stores, leading to a calcium flux within the T cell (Crabtree, 2001). This, in turn,
results in the Ca2+ mediated binding of calmodulin to calcineurin (Rusnak and Mertz,
2000), and the nuclear localization of the transcription factor NFAT (Rao et al., 1997).
DAG, on the other hand, activates protein kinase C, which leads to the phosphorylation
of IKB, allowing for the nuclear localization of transcription factor NF-KB (Tan and
Parker, 2003).
Furthermore, through the adaptor proteins Grb2 and Sos, LAT
phosphorylation leads to the activation of the Ras/Raf/MAP-kinase pathway, and results
in the nuclear localization and transcriptional activity of AP1 (Genot and Cantrell, 2000).
The TCR/MHC interaction is not sufficient to fully induce naive T cell activation.
In addition, a second, antigen nonspecific co-stimulatory signal is also required. This
second signal is usually provided by the interactions between T cell-expressed CD28
molecules and members of the B7 family, B7.1 (CD80) and B7.2 (CD86), which are
expressed on activated professional APCs (Acuto and Michel, 2003). Binding of CD28
by these molecules synergistically enhances TCR-induced proliferation and activation,
while TCR engagement to a peptide/MHC complex without co-stimulation can lead to T
cell anergy (Appleman and Boussiotis, 2003). This limits T cell activation to instances
13
when the innate immune system has already detected a pathogenic attack. This plays a
role in the prevention of T cell response to harmless environmental antigen, and helps
protect against self-reactive T cells that escaped negative selection.
In addition to CD28, the T cell expressed surface molecule CTLA-4 also is able to
interact with B7 family members. Naive T cells express CD28, but not CTLA-4. During
T cell activation, CTLA-4 expression is rapidly upregulated. While they share a ligand,
CTLA-4 and CD28 have antagonistic effects. While CD28 engagement enhances T cell
activation, CTLA-4 engagement inhibits activation (Alegre et al., 2001). As T cell
activation and CD28 engagement lead to increased expression of CTLA-4, CTLA-4 is
able to provide a braking mechanism that prevents an overly robust immune response.
This function is further evidenced by the fact that CTLA-4 deficient mice generate T cells
that proliferate excessively, leading to an engorged spleen and lymph nodes, and death
within 3 to 4 weeks (Tivol et al., 1995; Waterhouse et al., 1995).
HELPER T CELL POLARIZATION
In order to effectively and efficiently respond to various forms of pathogenic
infection, the mammalian immune system is able to detect the nature of the invading
pathogen and coordinate its response accordingly. One of the primary mechanisms used
to accomplish this targeted response is helper T cell polarization. Intracellular pathogens,
like virus or intracellular bacteria, induce a type 1 response (Romagnani, 1994; Spellberg
and Edwards, 2001).
During this response, antigen-specific naive helper T cells
differentiate into activated type 1 helper T cells (THi) and secrete interferon-y (IFN-y),
tumor necrosis factor (TNF)
and TNF3 (Jankovic et al., 2001; Mosmann et al., 1986;
Murphy and Reiner, 2002). Largely through this cytokine secretion, THI cells mediate
the activation of macrophages, the induction of IgG antibodies that mediate phagocytosis,
the support of cytotoxic CD8+ T cells, and further TH polarization (Boehm et al., 1997;
Spellberg and Edwards, 2001). A type 2 immune response, on the other hand, is induced
if the pathogenic infection consists of helminth, an extracellular intestinal parasite
(Pearce et al., 1991; Scott et al., 1989). In this case, naive helper T cells differentiate into
activated type 2 helper T cells (TH2 ) and begin to secrete a number of cytokines,
14
including interleukin(IL)-4, IL-5, IL-9, IL-10 and IL-13 (Jankovic et al., 2001; Mosmann
et al., 1986; Murphy and Reiner, 2002).
T2 cells support the expansion and
differentiation of eosinophils and mast cells, further TH2 polarization, and enhance the
production of IgE by B cells (Finkelman et al., 2004; Jankovic et al., 2001).
Consequences of Improper Helper T cell Polarization
The importance of proper helper T cell polarization is emphasized by the
consequences of improper polarization. One of the most straightforward potential effects
of improper polarization is an ineffective immune response. In mice, one example of this
is the response to Leishmania major, an intracellular parasite, by different mouse strains.
C57B16 mice are genetically predisposed towards type 1 polarization while the Balb/c
mice are genetically predisposed towards type 2 polarization (Himmelrich et al., 1998).
During a Leishmania infection C57B16 mice mount a TH1 response and rapidly clear the
pathogen, while Balb/c mice mount a TH2 response and are not protected from infection
(Heinzel et al., 1988; Sadick et al., 1986). This difference in pathogen susceptibility was
shown to be a direct effect of the helper T cell polarization when it was demonstrated that
blockade of IL-4 in Balb/c mice allowed them to mount a protective TH1 response
(Heinzel et al., 1991; Kopf et al., 1996). Conversely, when C57B16 mice were given IL-4
during the Leishmania infection they mounted a type 2 response, and were rendered
susceptible to pathogenic infection (Chatelain et al., 1999).
Improper helper T cell polarization has also been shown to determine the outcome
of a pathogenic infection has also been shown to occur in humans. A classic example of
this phenomenon is seen during human Mycobacterium leprae infections.
Mycobacterium leprae are the mycobacteria that cause leprosy. When a cell-mediated
1
H1 immune response is mounted in the infected host, the resulting disease is tuberculoid
leprosy, and most of the mycobacteria are effectively cleared, the infection remains
localized, and the host usually survives (Misra et al., 1995a; Misra et al., 1995b). On the
other hand, when a host responds to the pathogenic infection with a humoral TH2
response, the resulting disease is lepromatous leprosy (Misra et al., 1995b). In this case,
immune clearance is completely ineffective and the mycobacteria become widely
15
disseminated throughout the host. This results in extensive bone, cartilage and nerve
damage and the probable death of the host.
Improper balance of helper T cell polarization has also been implicated in
autoimmune disease and allergic disorders.
While the exact causes of organ specific
autoimmune diseases remain largely unknown, there is a strong correlation between these
syndromes and the production of TH cytokines. Among the diseases in which this has
been reported are multiple sclerosis (Brod et al., 1991) and Graves' disease (de Carli et
al., 1993). This correlation has also been shown in mouse models for organ specific
autoimmune disorders, including experimental immune uveoretinitis and experimental
allergic encephalomyelitis (Romagnani, 1994). Despite these correlations, causality has
yet to be convincingly demonstrated.
There is stronger evidence for a role of helper T cell polarization in allergies. One
of the major mediators of an allergic response is the expression of IgE antibodies by B
cells (Gould et al., 2003). For example, allergen-specific IgE has been shown to play a
major role in both allergic asthma and atopic dermatitis (Romagnani, 1994). T
2
cells
have an important role in this, as class switching to IgE is induced by the IL-4 (Gould et
al., 2003). Analysis of allergen-specific T lymphocytes show a strong predominance of
TH2 cells in individuals that show allergic symptoms, but a predominance of THl cells
from individuals that do not show allergic symptoms (Ebner et al., 1993; Wierenga et al.,
1990). That TH2 polarization plays a role in allergic asthma is the basis for the "hygiene
hypothesis," which states that the relatively high level of asthma in developed countries is
at least partially due to a lack of exposure to TH polarizing microorganisms during
childhood (Yazdanbakhsh et al., 2002). This results in largely TH2 polarized immune
system, and, consequently, a high prevalence of TH2 mediated disorders, like asthma.
While compelling, there is little experimental evidence that directly demonstrates the
validity of the hygiene hypothesis, and it is thus hotly debated.
I)endritic Cell Mediation of Helper T cell Polarization
The first step in helper T cell polarization involves the recognition of the type of
pathogen that is infecting the host. There is no evidence that helper T cells themselves
are able to directly identify the invading pathogen but, rather, it appears that T cells rely
16
on the signals from the innate immune system, especially dendritic cells, to direct their
polarization.
Dendritic cells are professional antigen presenting cells produced in the bone
marrow. In an immature state, these cells migrate to likely sites of pathogen entry and
monitor these sites for signs of infection (Ardavin, 2003). Pathogenic infection is
detected through a set of receptors that have evolved to recognize specific conserved
PAMPs (Janeway and Medzhitov, 2002). Examples of this type of receptor include the
Toll-like receptor family. The eleven described members of this family each recognize a
different conserved PAMP, and, upon ligation, lead to dendritic cell activation,
maturation, and the production of proinflamitory cytokines. In this way the dendritic cell
is able detect potential pathogenic infections and contribute to the local innate immune
response (Janeway and Medzhitov, 2002).
In addition to alerting the innate immune system to the presence of a potential
pathogen, dendritic cells also play a key role in the initial activation and direction of the
adaptive immune response (Janeway and Medzhitov, 2002). Being professional APCs,
dendritic cells present peptides acquired from the extracellular environment on class II
MHC (Watts, 2004). Because there is a relatively low level of surface expression of both
MHC II and costimulatiory molecules such as CD80 and CD86 in immature dendritic
cells, immature dendritic cells are not very efficient at activating helper T cells
(Banchereau et al., 2000). This changes upon PAMP-induced dendritic cell maturation.
During this process, there is an up-regulation of class II MHC and co-stimulatory
molecules, allowing the dendritic cell to more efficiently present any pathogen-derived
peptides and activate antigen specific helper T cells (Banchereau et al., 2000).
In
addition to these changes, activated dendritic cells also express a set of T cell polarizing
molecules, the composition of which varies depending upon the identity of the initial
PAMP that drove the cell's maturation (de Jong et al., 2002; Kalinski et al., 1999)(Figure
3). Thus, if the PAMP is a molecule that is commonly derived from virus or intracellular
bacteria, ligation of the receptor specific to that PAMP will likely induce expression of
molecules that drive a TH1 response, which more effectively clears these types of
pathogens. On the other hand, if the initial stimuli is a PAMP commonly derived from a
helminth, the dendritic cell will upregulate molecules that drive a TH2 response (de Jong
17
Bacteria, Viru
Naive helper
T Cell
OL,
lation
IL-4
IL-2
IFN-y
TH 1
Effector
Effector
IL-4
H
Effector
Figure 3. Some of the extracellular signals involved in helper T
cell polarization.
18
et al., 2002; Kalinski et al., 1999). Thus, dendritic cells make a large contribution to the
polarization decision.
As described above, type 1 helper T cells are induced in response to viral or
intracellular bacterial infection, and secrete IFN-y, TNFx and TNF3 in order to mediate
the immune response. While many of the molecular signals that induce THi polarization
are known, the exact series of interactions that cause this polarization to occur in vivo is
uncertain.
It is clear that IL-12 plays a prominent role in THI polarization. IL-12 is readily
secreted by dendritic cells upon stimulation by several different pathogen types, including
bacteria, viruses, and protozoa, or their products, such as LPS or double-stranded RNA
(Trinchieri, 2003). IL-12 is able to induce IFN-y expression in T cells and natural killer
(NK) cells (Trinchieri, 2003), and is able to efficiently drive naive helper T cells to
differentiate to a TH.1phenotype upon TCR stimulation in vitro (Murphy et al., 2000).
Furthermore, mice lacking a functional IL-12 receptor have a reduced capacity to mediate
a THI response, though THI polarization still occurs (Kaplan et al., 1996b).
In addition to IL-12, two other members of the IL-12 family, IL-23 and IL-27,
have been identified as having a role in THI polarization. Both IL-23 and IL-27 can be
secreted from dendritic cells, monocytes, and macrophages (de Jong et al., 2005). Like
IL-12, both of these molecules are capable of driving IFN-y expression in T cells and NK
cells (Trinchieri et al., 2003). While both IL-12 and IL-27 induce IFN-y expression in
naive helper T cells, IL-27 requires the presence of IL-12 or IL-18 for this effect (Takeda
et al., 2003). Despite this, mice deficient in IL-27 signaling show impaired early IFN-y
production when infected with Leishmania major, Listeria monocytogenes or BCG (Chen
et al., 2000; Yoshida et al., 2001). IL-23, on the other hand, is inefficient at inducing
IFN-y production in naive helper T cells, but is able to induce proliferation and IFN-y
expression in memory T cells (Oppmann et al., 2000). Despite recent efforts to better
understand their function, the precise roles played by the IL-12 family member cytokines
in THi polarization in vivo remains unclear.
The primary effector cytokine of the TH 1 response, IFN-y, also plays a prominent
role in helper T cell polarization (Murphy and Reiner, 2002).
IFN-y is produced by NK
cells, activated cytotoxic CD8+ T cells and TH1 cells, but is probably not directly
19
produced by dendritic cells (Kapsenberg, 2003). As one of the major targets of both NK
cells and cytotoxic T cells is virally infected cells, it is not surprising that they may play a
role in the THI polarization
of the immune response.
That TH cells secrete a cytokine
that furthers type 1 polarization is also not surprising. As described below, this is only
one of several positive feedback loops that are involved in both type 1 and type 2
polarization. Finally, it should be noted that one of the most potent inducers of IFN-y
expression in both NK cells and T cells is IL-12 (Trinchieri et al., 2003), and thus IFN-y
production is often downstream of IL-12 signals.
Another cytokine that has been implicated in TH polarization is IL-18. Dendritic
cells, monocytes and macrophages constitutively express pro-IL-18. Upon exposure to
LPS or gram-positive bacteria, these cells also produce caspase-1, which cleaves pro-IL18 into biologically active IL-18 (Biet et al., 2002). While IL-18 alone induces low
levels of IFN-y in T cells, IL-18 enhances the IFN-y inducing activity of IL-12
(Barbulescu et al., 1998). This may occur in part because IL-12 induces T cell expression
of the IL-18 receptor, so in the absence of IL-12 T cells may be unable to detect the IL-18
(Munder et al., 1998).
The signals that drive naive helper T cells to differentiate into TH2 cells in vivo is
less clear than those involved in TH polarization. One of the most potent molecules for
inducing TH2 polarization is the type 2 effector cytokine IL-4. IL-4 is capable of driving
naive helper T cells to differentiate into effector TH2 cells upon in vitro activation (Scott
et al., 1989), but dendritic cells and other professional antigen presenting cells do not
normally express IL-4. In vivo, a significant source of IL-4 is the helper T cell population
itself, so it is likely IL-4 acts partly in an autocrine fashion, enhancing the effect of other
TH2 polarization stimuli.
There are dendritic cell-expressed molecules that have been suggested to play a
role in TH2 polarization, although the importance and effect of these molecules on helper
T cell polarization is unclear. Monocyte chemoattractant protein-I (MCP- 1)/CCL2 is a
chemochine that attracts monocytes, memory T cells and NK cells (Chensue et al., 1996).
MCP-1/CCL2 is expressed by mature dendritic cells, and in allergic and infectious
disease models it has been shown to enhance IL-4 production (Chensue et al., 1996).
This result is contradicted by the observation that mice deficient in CCR2, MCP-
20
I/CCL2's only known ligand, show a reduction in TH1 polarization (Boring et al., 1997).
Thus, the exact role of MCP- /CCL2 in helper T cell polarization remains unclear.
The dendritic cell expressed type II transmembrane molecule OX40L has also
been implicated as an inducer of TH2 polarization.
It has been demonstrated that
activation of naive helper T cells with an OX40L transfected cell line increases IL-4
secretion and enhances TH2 polarization in vitro (Flynn et al., 1998). Furthermore, it has
been demonstrated that dendritic cells that have been primed with IFN-y or poly I:C, both
of which induce THI polarization, fail to express OX40L, while dendritic cells primed
with PGE2 or schistosomal egg antigen, both of which induce TH2 polarization, express
detectable levels of OX40L, and that levels are further increased in these cells upon
CD40 ligation (de Jong et al., 2002). Confusingly, while the elevated OX40L levels
induced by the schistosomal egg antigen promoted a TH2 response, the increased levels
induced by PGE2 did not (de Jong et al., 2002), indicating that, while OX40L may play a
role in TH2 polarization, it is not sufficient to induce T.2 polarization.
While several molecules have been suggested to play a role in the initiation of this
polarization process, the strength of the activation signal that the T cell receives may also
contribute to this process. In the absence of other TH polarizing signals, there is a
correlation between a strong activation signal and TH2 polarization (Hosken et al., 1995;
Rulifson et al., 1997).
This strong signal can come directly through the TCR-
peptide/MHC interaction, where strong signals induce TH2 polarization, while weaker
signals tend to induce THi polarization (Hosken et al., 1995), or it can come through the
co-stimulatory molecule, as increased CD28 stimulation induces TH2 polarization
(Rulifson et al., 1997), and ligation of CTLA4, which inhibits T cell activation, promotes
TH1 polarization (Kato and Nariuchi,
2000). Indeed, T cells deficient in CTLA4
expression are especially prone to IL-4 independent TH2 polarization (Bour-Jordan et al.,
2003).
Whether this correlation between TCR signal strength and helper T cell
polarization plays a significant role in vivo is unknown.
IL-2 may also contribute to the initiation of TH2 polarization.
It has been
demonstrated that recently activated helper T cells have reduced capacity of expressing
IL-4 if they are cultured in the presence of IL-2 neutralizing antibodies (Zhu et al., 2003).
IL_-2receptor ligation induces the activation of StatS, and it has been demonstrated that T
21
cells containing constitutively active Stat5 express IL-4 even when polarized under THi
conditions (Zhu et al., 2003). Thus, it is likely naive helper T cells that are activated
without THI polarizing cytokines present, but in the presence of IL-2 are able to initiate
expression of IL-4. IL-4 would then be able to act as an autocrine factor, enhancing the
polarization of the secreting cell, as well as driving TH2 polarization in other activated but
undifferentiated helper T cells.
suggested to induce
TH2
Exactly how the other molecules that have been
polarization contribute to this process is unclear, and may vary
depending upon unknown conditions.
Molecular Regulation of Helper T cell Polarization
As discussed above, naive helper T cells can differentiate into at least two
phenotypes depending upon the conditions of activation. Under conditions induced by
intracellular pathogens, polarization to a TH phenotype is favored, while under
conditions induced by helminths, polarization to a TH2 phenotype is favored.
The
primary distinction between these phenotypes lies in the cytokines they express. The
effector cytokine of TH1 cells is IFN-y, while the effector cytokine of TH2cells is IL-4. In
addition to being secreted by THI and T2 cells respectively, IFN-y and IL-4 also
contribute to helper T cells polarization. Exposure to IFN-y enhances TH1 polarization
and inhibits TH2 polarization, while exposure to IL-4 enhances TH2 polarization and
inhibits TH1 polarization.
These opposing positive feedback loops help ensure that
polarization of helper T cells is a permanent and complete differentiation event.
When naive helper T cells become activated, they can polarize into either a TH1
phenotype or a T,,2 phenotype. This polarization process requires both of the T cell
activation signals, but the direction of polarization is determined by the context of the
activation (essentially a third signal).
While this third signal is required for full
polarization, it is not required for the initiation of transcription of either IL-4 or IFN-y.
The engagement of the TCR complex with an agonist peptide/MHC II complex results in
a CD3 mediated signal cascade that, among other things, induces the nuclear localization
of AP1 and NFAT family members. The presence of these important transcription
factors in the nucleus contributes an early, low level, polarization independent
transcription of IL-2, IFN-y, and IL-4 (Grogan et al., 2001). It is important to note that
22
the expression of both effector cytokines occurs within one hour of TCR engagement, far
faster than the upregulation of the THI and TH2 master regulator genes T-bet and GATA3 respectively (Grogan et al., 2001). The signal provided through the TCR complex is
required for both TH 1 and TH2 polarization.
Helper T cell polarization is a complex and tightly regulated differentiation event.
The process by which CD4+ T cells convert extracellular signals into polarization-specific
gene expression involves many different interacting molecules and signal cascades.
Figure 4 is a schematic outlining some of the more important and better understood of
these interactions.
It is likely that, in vivo, the major signal for the initiation of THI polarization is
dendritic cell secreted IL-12. When a naive helper T cell becomes activated by an IL-12
secreting dendritic cell, IL-12 engagement results in the dimerization of the IL-12
receptor subunits, which in turn leads to the activation of the associated Janus kinases
Jak2 and Tyk-2 by reciprocal phosphorylation (Watford et al., 2004). The Janus kinases
are then able to phosphorylate Stat4, leading to its dimerization and translocation to the
nucleus (Jacobson et al., 1995; Wurster et al., 2000). Nuclear localization of activated
Stat4 during helper T cell activation leads to an increased level of IFN-y transcription
(Watford et al., 2004). This is supported by the observation that Stat4 deficient mice are
defective in IFN-y production (Magram et al., 1996). While there are several possible
Stat4 binding sites present within the IFN-y promoter, it is unclear whether Stat4 directly
interacts with this locus, and thus the IFN-y induction may be an indirect effect.
Regardless, through Stat4, the IL-12 receptor signal leads to increased transcription of the
TH 1 effector cytokine IFN-y.
While IFN-y alone does not fully induce TH1 polarization, it does play a role in
the process (Szabo et al., 1997; Wenner et al., 1996) and is a potent repressor of TH2
polarization (Murphy and Reiner, 2002). Because of this, the IFN-y induced by the IL-12
signal can act both as an autocrine signal, and as an inducer of THi polarization in other,
undifferentiated helper T cells. In addition to TH1 cells, activated natural killer cells and
activated CD8+ cytotoxic T cells also secrete significant levels of IFN-y, and can thus
contribute to the TH 1 polarization of helper T cells. IFN-y facilitates the dimerization of
its receptor, which leads to the reciprocal phosphorylation of the associated Janus
23
\
-
IL-4
Receptor
Receptor
4,a
Figure 4. Some of the molecular interactions important
for helper T cell polarization.
24
kinases, Jakl and Jak2 (Pestka et al., 1997). This leads to the phosphorylation,
dimerization, and nuclear localization of Statl (Pestka et al., 1997). Statl deficient T
cells show impairment in IFN-y expression and IL-12R32 upregulation following TH
polarization (Afkarian et al., 2002). Both of these phenotypes can be explained because
IFN-y receptor signaling through Statl activation induces the transcription of the master
regulator ofTH polarization, T-bet (Afkarian et al., 2002; Lighvani et al., 2001).
T-bet's critical role in TH1 polarization is supported by the fact that helper T cells
that lack functional T bet are unable to efficiently polarize to a TH1 phenotype (Szabo et
al., 2002), and that ectopic T-bet expression is able to inhibit TH2 cytokines and induce
IFN-y expression in differentiated TH2 cells (Szabo et al., 2000). While much is still
being learned about T-bet's role in THI polarization, several of T-bet's functions that
contribute to this process have been elucidated. Among these functions, T-bet has been
shown to be important for the opening of the IFN-y locus (Mullen et al., 2001) and the
induction of acute IFN-y transcription (Szabo et al., 2000), both through a direct
interaction with the IFN-y promoter (Tong et al., 2005) and by upregulating the IFN-y
inducing transcription factor Hlx (Mullen et al., 2002). T-bet has also been shown to
upregulate transcription of the IL-12 receptor
(Mullen et al., 2001), as well as its own
transcription (Mullen et al., 2002), further strengthening the TH polarization signal. In
addition, T-bet has been shown to suppress the expression of IL-2 through a direct
interaction with the IL-2 promoter (Szabo et al., 2000) (Hwang et al., 2005a). It has also
been demonstrated that T-bet undergoes a Itk mediated phosphorylation at tyrosine 525
that enables it to inhibit the function of the TH2 master regulator, GATA-3 through a
direct interaction with the GATA-3 protein (Hwang et al., 2005b). Thus, T-bet is able to
both enhance TH polarization by enhancing expression of IFN-y and IL12R3, and
repress TH2polarization by inhibiting GATA-3.
The best understood inducer of TH2 polarization is the cytokine IL-4. The IL-4
receptor signal, like the previously described IL-12 and IFN-y receptor signals, is
mediated through a Jak/Stat interaction.
Briefly, binding of IL-4 mediates the
dimerization of the IL-4 receptor ccchain with the common gamma chain, which, in turn,
leads to the mutual phosphorylation and activation of the associated Janus tyrosine
kinases, Jakl and Jak3 (Nelms et al., 1999). Jak mediated phosphorylation of specific
25
tyrosine residues on the IL-4 receptor a chain, allows Stat6 to bind through its Src
homology
2 domain (Nelms et al., 1999).
This interaction results in Stat6
phosphorylation by the Janus kinases, leading to Stat6 dimerization and translocation to
the nucleus (Wurster et al., 2000).
Stat6 plays an important role in TH2 polarization, illustrated by the fact that TH2
deficiency that is a major phenotype of Stat6 deficient mice (Kaplan et al., 1996a; Takeda
et al., 1996). Stat6 activation has been shown to induce the expression of the TH2 master
regulator, GATA-3 (Ouyang et al., 1998). While other downstream Stat6 targets have
been suggested, much of Stat6's TH2 polarizing activity can be attributed to this induction
of GATA-3 expression. This is evidenced by the fact that TH2 responses can be rescued
in Stat6 deficient mice through ectopic GATA-3 expression by T cells (Ouyang et al.,
2000).
GATA-3 is highly expressed in TH2cells, but not in TH1 cells, and plays a central
role in TH2 polarization (Zhang et al., 1997). GATA-3 deficiency is embryonic lethal
(Pandolfi et al., 1995), and deficiency in the lymphoid system results in a complete block
in early T cell development (Hendriks et al., 1999), illustrating that GATA-3 plays an
important role in many developmental processes. Disruption of GATA-3 in mature T
cells results in an inhibition of TH2 polarization (Pai et al., 2004; Zhu et al., 2004), and
expression of GATA-3 is sufficient to induce TH2 polarization in naive helper T cells
(Lee et al., 2000).
High levels of ectopic GATA-3 expression in TH1 cells is sufficient to
induce the expression of TH2 cytokines while inhibiting the expression of TH1 cytokines
(Lee et al., 2000; Zhang et al., 1997).
As can been seen in Figure 4, GATA-3 exerts its TH2 polarizing function through
several mechanisms. In addition to enhancing the its own expression (Ouyang et al.,
2000), GATA-3 has been shown to promote the expression of c-Maf (Ouyang et al.,
2000), which in turn enhances the expression of IL-4 through a direct interaction with the
IL-4 proximal promoter (Ho et al., 1996; Lee et al., 2000). GATA-3 has also been shown
to contribute to the expression of the TH2 cytokine genes by promoting the epigenetic
"opening" of the TH2 cytokine gene cluster (Ouyang et al., 2000; Takemoto et al., 2000).
While there is little evidence to suggest that GATA-3 interacts directly with the IL-4
promoter, it is able to drive expression of IL-5 and IL-13 through direct interactions with
26
locus regulatory regions (Kishikawa et al., 2001; Zhang et al., 1997). In addition to its
positive effect on the expression of TH2 cytokines, GATA-3 has also been shown to
inhibit T
polarization.
While the exact mechanisms of this inhibition are not
completely clear, GATA-3 has been shown to inhibit the ability of Stat4 to enhance IFNy expression (Usui et al., 2003), and to prevent the upregulation of the IL-12 receptor
(Ouyang et al., 1998).
The initial molecular interactions in an in vivo TH2 response are not well
understood. While it is clear that IL-4 can efficiently induce TH2 polarization in vitro,
there is little evidence that dendritic cells, or other professional antigen presenting cells,
are capable of expressing IL-4. The major source of IL-4 in mammals is, in fact,
activated TH2 cells (de Jong et al., 2005), leading to the possibility that the major roles of
IL-4 in TH2 polarization are by acting as an autocrine signal, and by contributing to the
further TH2 polarization of helper T cells. It is therefore possible that the initial stages of
TH2 polarization are IL-4 independent.
The hypothesis of IL-4 independent TH2 polarization is strengthened by the
observation that Stat6 deficient mice are still able to mount a limited TH2 response
(Ouyang et al., 2000), and TH2 cells from these mice express normal levels of TH2
cytokines and high levels of GATA-3 (Ouyang et al., 2000). So the question arises, how
can TH2 polarization occur in the absence of IL-4 and Stat6?
While there are several molecules that could contribute to Stat6 independent TH2
polarization, it is possible that a strong activation signal in the absence of TH polarizing
signals could initiate this polarization process. Nafve helper T cells have a low level of
GATA-3 expression that gets upregulated under TH2 polarizing conditions (Grogan et al.,
2001). It has been shown that this low level of GATA-3, in conjunction with a strong
TCR signal, is sufficient to induce IL-4 transcription (Grogan et al., 2001; Zhu et al.,
2004). Supporting this, IL-4 transcription can be detected within one hour of TCR
stimulation under both TH1 and TH2 polarizing conditions, and GATA-3 upregulation is
not seen until 24 to 72 hours after TCR stimulation and only under TH2 polarizing
conditions (Grogan and Locksley, 2002; Hwang et al., 2005b). Furthermore, it has been
shown that a strong activation signal from the antigen presenting cell favors TH2
polarization (Hosken et al., 1995; Rulifson et al., 1997). This could be why helper T cells
27
deficient in the inhibitory B7 ligand CTLA-4 are skewed towards T 112 differentiation,
even in the absence of Stat6 (Kato and Nariuchi, 2000), and why high amounts of CD28
costimulation also skews helper T cells towards TH2 differentiation (Rulifson et al.,
1997). Thus, it is possible that a strong activation signal in the absence of THI polarizing
signals could induce the initial expression of IL-4, which could then act in an autocrine
fashion through Stat6 to re-enforce the polarization. Once the TH2 response was initiated,
the early TH2 cells could provide the IL-4 signal to push undifferentiated helper T cells
towards TH2 polarization.
In addition to IL-4, it is likely that IL-2 plays a role in promoting a TH2 response.
IL-2 is rapidly produced by all newly activated T cells independent of polarizing
conditions (Jain et al., 1995), so autocrine IL-2 will likely be available during the initial
stages of a TH2 response. It has been shown that IL-2 neutralizing antibodies inhibit TH2
polarization (Zhu et al., 2003), suggesting that IL-2 signaling enhances TH2 polarization.
Furthermore, the downstream molecular mediator of the IL-2 receptor is Stat 5, and
activated helper T cells that express constitutively active Stat5 express IL-4, even when
activated under THI polarizing conditions (Zhu et al., 2003), while helper T cell
polarization is biased towards THI in Stat5 -' - mice (Kagami et al., 2001). Thus, IL-2
autocrine signaling could contribute to the effect of a strong activation signal in the
absence of polarizing cytokines to initiate TH2 polarization.
REGULATION OF IFN-y TRANSCRIPTION BY PROXIMAL PROMOTER
BINDING FACTORS
In addition to being the effector cytokine of TH1 cells, IFN-y is also expressed by
natural killer cells and activated cytotoxic CD8 T cells. IFN-y is involved in the
activation of macrophages, antiviral immunity, and protection against tumorogenesis
(Dredge et al., 2002; Romagnani, 1994; Spellberg and Edwards, 2001).
Given its
importance in immunity and its wide range of effects, it is not surprising that much work
has gone into examining the cis-elements and transcription factors that regulate IFN-y
expression.
28
Figure 5 is a schematic of the IFN-y locus indicating some of the known and
suspected transcription factor binding sites. Included among the factors that have been
demonstrated to interact with the IFN-y proximal promoter are TH specific factors (Stat4
and T-bet), factors induced by the TCR signal (AP-1 and NFAT), and ubiquitously
expressed factors, like Ying-Yang 1.
One of the most important transcription
differentiation., but not during TH2 differentiation
factors upregulated during TH
is T-bet.
T-bet expression is
downstream of the IFN-y receptor and Statl activation (Afkarian et al., 2002). While Tbet has been shown to have an important role in the epigenetic opening of the IFN-y locus
(Mullen et al., 2001), its role in driving transcription is less clear. T-bet overexpression
has been shown to enhance IFN-y proximal promoter driven reporter gene expression
(Soutto et al., 2002), and mutation of a conserved T-box half site located 63 bp upstream
of the transcription start site inhibits this enhancement (Tong et al., 2005). Furthermore,
chromatin immunoprecipitation (ChIP) assays have indicated that T-bet is able to bind to
the IFN-y proximal promoter following PMA and ionomycin stimulation (Tong et al.,
2005).
While the above experiments indicate that T-bet is able to enhance IFN-y
proximal promoter function, how much it does so in vivo is less clear. It has been
demonstrated that, independent of polarizing conditions, IFN-y transcription is initiated
within 1 hour of T cell stimulation, and peaks within 3 hours (Grogan et al., 2001). T-bet
transcription, on the other hand, is not upregulated for 24 to 48 hours after activation
(Groganet al., 2001; Hwang et al., 2005b), and then only under TH1 polarizing conditions
(Grogan et al., 2001; Szabo et al., 2000). This indicates that early IFN-y transcription is
T-bet independent. Also, while CD4+ T cells and NK cells from T-bet deficient mice
show a defect in IFN-y expression, CD8+ cells from these same mice express normal
levels of T-bet (Szabo et al., 2002). Thus, while T-bet may play a direct role in IFN-y
transcription in TH1 cells and NK cells, it is not necessary for IFN-y expression in
cytotoxic T cells.
Finally, expression of a dominant negative T-bet protein is able to
block IFN-y expression and THI polarization in newly activated helper T cells, but is
unable to block IFN-y expression in fully differentiated TH1 cells (Martins et al., 2005;
Mullen et al., 2002). This reinforces the idea that T-bet's main role in IFN-y expression
29
The IFN-y Proximal Promoter
Stat4
YY1
YY1 AP1
-240
-223
-205 -200
NFAT AP1
-98
-85
T-bet AP1
-63
-53
If.
0
Figure 5. A schematic representation of the IFN-y proximal
promoter indicating some of the transcription factor binding
sites.
30
is the epigenetic opening of the locus, rather than acting as a direct mediator of
transcription.
The engagement of the IL-12 receptor with its ligand leads to the activation and
nuclear localization of Stat4 (Jacobson et al., 1995; Wurster et al., 2000). Stat4 has been
shown to play a critical role in TH1 polarization, potentially through a direct interaction
with the IFN-y promoter (Xu et al., 1996). In transient transfection assays it has been
shown that Stat4 is able to enhance IFN-y proximal promoter driven luciferase expression
(Barbulescu et al., 1998). While it has been shown by EMSA that Stat4 is able to bind to
a sequence located 240 bp upstream of the transcription start site of the human IFN-y
proximal promoter (Barbulescu et al., 1998), this sequence is not conserved in mice and,
when the analogous mouse sequence was used in a similar EMSA, no binding was
detected (Nakahira et al., 2002). This could indicate that Stat4 plays a different role in
the regulation of the IFN-y proximal promoter in humans than in mice, or it could
indicate that the in vitro binding (or lack thereof) demonstrated in these assays is not
representative of what happens in vivo. Therefore, while it is clear that Stat4 is important
for IFN-y proximal promoter function, it is not certain whether Stat4 mediates its activity
through a direct interaction with the promoter.
Yin-Yang 1 (YY1) is a zinc-finger transcription factor that is expressed
ubiquitously within the nucleus (Shi et al., 1991), and has been shown to have both
activating and repressing properties (Thomas and Seto, 1999). Two YY 1 binding sites
have been demonstrated in the IFN-y proximal promoter, one at -205 and the other at
*-223 relative to the transcription start site (Ye et al., 1996). YY 1 has been shown to have
a repressive effect when bound to these sites. Mutation of either of the YY 1 binding sites
enhances promoter function in transient transfection assays (Ye et al., 1996). There is
evidence that suggests that one mechanism by which YY1 inhibits IFN-y proximal
promoter function is by blocking AP1 factors from binding to a site just downstream of
the -205 YY1 site (Ye et al., 1996). Furthermore, YY 1 has been shown to play a role in
the chromatin remodeling, and specifically the recruitment of histone deacetylases (Yang
et al., 1996; Yang et al., 1997). While there is little evidence either for or against it
functioning in this way at the IFN-y promoter, changes in the histone acetylation pattern
have been shown to play a role in IFN-y regulation (Avni et al., 2002). It is therefore
31
possible that one mechanism that YY 1 uses for suppression of IFN-y transcription is
chromatin modification.
In T cells, the NFAT family of transcription factors becomes dephosphorylated by
the calmodulin-dependent phosphatase calcineurin downstream of the calcium flux that is
initiated by TCR engagement (Macian, 2005). This dephosphorylation leads to NFAT
nuclear translocation, where the NFAT proteins contribute to the transcription of a wide
range of genes involved in the immune response, including IFN-y (Macian, 2005). Mice
deficient in NFAT1 or deficient in both NFAT1 and NFAT4 show a skewing towards
TH2
polarization, and a decrease in IFN-y expression (Kiani et al., 2001; Rengarajan et
al., 2002), and NFAT binding to the IFN-y promoter in TH1, but not TH2 cells has been
demonstrated by ChIP assays (Agarwal et al., 2000).
Notably, NFAT often acts
synergistically with the AP1 family of transcription factors (Macian et al., 2001), and a
probable NFAT binding site is located immediately upstream of an AP1 binding site in
the IFN-y promoter.
Thus, it is likely NFAT/AP1 cooperation occurs at the IFN-y locus.
As NFAT has also been shown to play an important role in IFN-y transcription by CD8+
T cells (Teixeira et al., 2005), it is likely that NFAT acts as a general inducer of IFN-y
expression downstream of the TCR signal.
AP1, like NFAT, has been acts downstream of the TCR and CD28 signals (Genot
and Cantrell, 2000). The AP1 family acts as heterodimers or homodimers of Jun and Fos
transcription factors (Hess et al., 2004).
Complicating matters, members of the
CREB/ATF family of transcription factors often are able to bind to the same sites as AP1
family members, and can even form heterodimers with AP1 family members (Hai and
Curran, 1991; Hess et al., 2004). The role of AP1 family members on the transcriptional
regulation of the IFN-y promoter is further complicated by the fact that there are at least
three AP1 binding sites within the proximal promoter. The most distal of the sites,
located 200 bp upstream of the transcription start site, has been shown to bind c-Fos and
c-Jun in vitro (Ye et al., 1996). As described above, it has been indicated that AP1
family member binding to this site is inhibited by the presence of YY 1 (Ye et al., 1996).
A second AP1 binding site is located between -98 and -78 bp relative to the IFN-y
transcription start site (Penix et al., 1993; Zhang et al., 1998a). A dimer of this site is
sufficient to drive reporter gene expression in IFN-y expressing cells (Penix et al., 1993),
32
and its activity is significantly enhanced by IL-12 (Zhang et al., 1998a). Members of the
Jun and ATF families of transcription factors have been shown to bind to this site in vitro
(Zhang et al., 1998a). The most proximal of the AP1 binding sites in the IFN-y proximal
promoter is located between -70 and -44 bp relative to the IFN-y transcription start site
(Penix et al., 1993; Penix et al., 1996). Like the site located between -98 and -78 bp,
multiple members of the Jun, ATF and CREB families have been shown to bind this site
in vitro (Penix et al., 1996), and multimers of this site are sufficient to drive reporter gene
expression in IFN-y expressing cells (Penix et al., 1996). Interestingly, a CpG located
within this site has been shown to be methylated in TH2 cell lines, but not in TH2 cell lines
(Young et al., 1994), and methylation of this CpG has been shown to affect factor binding
in vitro (Young et al., 1994), though it is unclear what factors are involved in this effect.
There is evidence that differential expression of the AP1 and CREB/ATF family
members plays a role in IFN-y expression.
It has been demonstrated that a dominant
negative mutant of c-Jun inhibits the function of the entire IFN-y proximal promoter,
multimers of the AP1 binding site located between -98 and -78 bp, or multimers of the
site located between -70 and -44 bp relative to the transcription start site (Cippitelli et
al., 1995; Penix et al., 1996). This indicates that c-Jun is likely to play an activating role
in IFN-y expression.
Likewise, overexpression of CREB inhibits the function of the
entire IFN-y proximal promoter, multimers of the AP1 binding site located between -98
and -78 bp, or multimers of the site located between -70 and -44 bp relative to the
transcription start site (Penix et al., 1996; Zhang et al., 1998a). This indicates that CREB
is likely to play an inhibiting role in IFN-y expression.
Furthermore,
in naive helper T
cells, there is a relatively high nuclear expression level of both CREB and c-Jun, but
upon TCR stimulation and the initiation of IFN-y expression, the level of CREB
diminishes, while the level of c-Jun remains constant (Zhang et al., 1998b). This
suggests that c-Jun and CREB, which may both be able to bind two of the AP1 binding
sites in the IFN-y proximal promoter, may have mutually antagonistic effects with CREB
inhibiting IFN-y transcription and c-Jun enhancing transcription. CREB binding to the
AP1 sites in nai've cells may prevent IFN-y expression.
Following TCR stimulation,
nuclear CREB levels drop, which may allow c-Jun access to the IFN-y proximal
33
promoter, initiating early transcription. While consistent with current observations, this
model has yet to be verified.
EPIGENETIC REGULATION OF HELPER T CELL POLARIZATION
The signaling cascades that lead a newly activated helper T cell differentiate into
either a THl cell or a TH2 cell have many effects. In addition to changing the nuclear
transcription factor composition, helper T cells polarization also changes how different
cytokine genes are packaged (Ansel et al., 2003; Grogan and Locksley, 2002). Through a
combination of DNA methylation, histone modification, and nuclear localization, a
polarized helper T cell enhances access to one set of cytokines while restricting access to
another. The heritable nature of these changes is thought to ensure long-term gene
silencing of following polarization.
DNA Methylation
The methylation of the 5 position of cytosine residues in CpG dinucleotides has
been shown to play a role in the regulation of mammalian genes (Jeltsch, 2002; Richards
and Elgin, 2002). This nucleotide modification does not interfere with Watson-Crick
basepairing, but is located within the major groove of the DNA double-helix, and thus
can be readily recognized by interacting transcription factors. While this modification
has been shown to occur at between 60-90% of CpG dinucleotides in mammals (Jeltsch,
2002), it is almost completely absent from non-CpG cytosines. As both strands of a CpG
palendromic motif generally share the same methylation state (either methylated or
unmethylated), DNA replication results in the generation of hemi-methylated DNA. A
DNA methyltransferase specific for hemi-methylated DNA can then methylate the hemimethylated daughter strands, allowing for the maintenance of the DNA methylation
pattern following cell division (Stein et al., 1982).
This property makes DNA
methylation a useful mechanism for long-term gene silencing following cell
differentiation.
There is a strong correlation between DNA methylation and gene transcription,
whereby silent genes are often hypermethylated, and transcribed genes are usually
34
hypomethylated (Siegfried and Cedar, 1997). This suggests that CpG methylation plays a
role in the repression of transcription.
There are three mechanisms by which this
covalent DNA modification may exert this effect, through the direct inhibition of
transcription factor binding, through the recruitment of methylated CpG binding proteins
that directly inhibit transcription, and through the recruitment of factors that influence the
way the DNA is packaged, and in doing so make the DNA less accessible to transcription
factors.
While some transcription factors show no preference for methylated DNA versus
unmethylated DNA, others have shown a decreased affinity for certain methylated
sequences versus their unmethylated counterparts.
Included among these genes are
several involved in immune function, including AP1, AP2, and NF-KB (Siegfried and
Cedar, 1997). It is therefore possible that the function of such transcription factors could
be regulated by DNA methylation.
Indeed, direct inhibition of transcription factor
binding by methylated DNA at several genetic loci has been well demonstrated in vitro
by EMSA (Dong et al., 2000; Fujimoto et al., 2005; Young et al., 1994). Included among
the sites where methylation has been shown to affect factor binding is the IFN-y promoter
(Young et al., 1994). Thus, while it is likely that the direct inhibition of factor binding by
DNA methylation has some effect on gene expression, it remains unclear how important
this mechanism is for transcription inhibition in vivo.
Several factors able to mediate transcriptional inhibition have been found that
bind preferentially to methylated DNA, including MeCP2, MBD1 and MBD2 (Cross et
al., 1997; Hendrich and Bird, 1998). Given these factors' ability to bind to methylated
CpGs, it is possible that they could act as repressors by simply blocking access of
activating transcription factors to hypermethylated promoters. Furthermore, the MeCP2
complex has been shown to have a repressor domain that can inhibit promoter
transcription at a distance (Kalinski et al., 1999), suggesting that MeCP2 can interact with
transcriptional machinery, in addition to blocking specific transcription factors.
In addition to directly acting as transcription repressors, methyl-CpG binding
proteins have also been implicated in the induction of further chromatin remodeling
through the recruitment of histone modification proteins to hypermethylated promoters
(Fujita et al., 2003; Fuks et al., 2003; Jones et al., 1998; Nan et al., 1998). Modifications
35
to the amino terminal histone tails, including methylation, acetylation, phosphorylation
and ubiquitination have been shown to play a role in gene accessibility (see below).
Histone acetylation of histones H3 and H4 is associated with transcriptionally active
genes, where a lack of histone acetylation is associated with silent genes (Richards and
Elgin, 2002). It has been demonstrated that MeCP2, MBD1 and MBD2 can all recruit
histone deacetylaces to regions with methylated DNA (Jones et al., 1998; Nan et al.,
1998). Through these factors, DNA methylation can induce the local hypoacetylation of
a locus, reducing the accessibility of the site. This effect can be further strengthened
because methyl-CpG binding proteins have also been demonstrated to recruit histone
methyltransferases to regions of hypermethylated DNA (Fujita et al., 2003; Fuks et al.,
2003). As methylation of lysines on histone H3 have been shown to play a role in
transcription regulation (Richards and Elgin, 2002), this histone methyltransferase
recruitment may also play a role in transcriptional suppression.
In order to function in gene transcription regulation, there must me mechanisms in
place to establish and maintain methylation patterns. Enzymes involved in both de novo
CpG methylation and in the maintenance of methylation during DNA replication have
been identified and their activities established in vitro. However, very little is known
about how these methyltransferases are targeted to a specific locus, or even which
enzymes methylated specific genes.
Furthermore, while there is evidence for the
existence of an active demethylation mechanism, no active demethylase has been
convincingly identified.
Then first mammalian methyltransferase
identified, Dnmt 1, preferentially
methylates hemimethylated DNA compared to unmethylated DNA (Bestor et al., 1988;
Bestor, 1992). Dmntl has been shown to be expressed ubiquitously in dividing cells, and
to localize to the replication fork (Leonhardt et al., 1992). It is therefore likely that
Dnmtl acts primarily as a maintenance methyltranferase, maintaining the established
methylation patterns during DNA replication.
It has been shown that dnmtl can
methylate unmethylated DNA in vitro (Bestor, 1992), but there is little in vivo evidence
that it functions as a de novo methyltransferase.
Mice lacking Dnmtl die during
embryogenesis (Li et al., 1992) at which point they show a dramatic reduction in global
36
methylation patterns (Li et al., 1992), indicating the importance of CpG methylation
during development.
Both in vitro and in vivo, the methyltransferase dnmt3a shows an enzymatic
preference for unmethylated DNA over hemimethylated DNA (Okano et al., 1998),
suggesting that it functions as de novo methyltransferase. Dnmt3a is highly expressed in
undifferentiated embryonic stem cells (Okano et al., 1998) and has broad somatic
expression across many tissues (Chen et al., 2002). Mice deficient in dnmt3a appear
normal at birth, but soon become runted and usually die within 4 weeks of birth (Okano
et al., 1999). Despite surviving to term, dnmt3a -' - embryos show reduced methylation
compared to wild-type embryos, suggesting that dnmt3a plays a role in the establishment
of methylation during early development (Okano et al., 1999). The role of dnmt3a in
somatic tissues is unknown.
Dnmt3a is partly regulated through alternative splicing and alternative
transcription start sites. There have been at least 5 separate splice variants of the dnmt3a
gene identified (Weisenberger et al., 2002).
As they all contain an intact enzymatic
domain, and it is likely that the proteins coded by these variants all have a level of
enzymatic activity, though this has yet to be established. The only reported difference
between these splice forms is in their expression pattern, with transcripts containing exon
Ia being preferentially expressed in somatic cells, and transcripts containing exon 113
being preferentially expressed in embryonic stem (ES) cells (Weisenberger et al., 2002).
Dnmt3a also has a functional intronic promoter, and transcription from this promoter
produces the transcriptional variant dnmt3a2 (Chen et al., 2002). Dnmt3a2 contains most
of Dnmt3al, including the enzymatic regions, but lack dnmt3al's 219 most N-terminal
amino acids (Chen et al., 2002). There is a dnmt3a2 specific intron, but it is upstream of
the translation start site, and thus does not code for protein. While both dnmt3al and
dnmt3a2 are expressed in ES cells and dnmt3a2 shows similar methylation activity in
vitro to dnmt3al (Chen et al., 2002), mice deficient in dnmt3al alone have a similar
phenotype to mice completely deficient for dnmt3a, but dnmt3a2 deficient mice appear to
be normal (personal communication).
It is also interesting that dnmt3al localizes to
heterochromatic regions of the nucleus, but dnmt3a2 localizes to euchromatic regions of
the nucleus, and where dnmt3al is expressed at similar levels across a broad array of
37
somatic tissues, dnmt3a2 expression appears to be largely restricted to the spleen, thymus
and testis (Chen et al., 2002). Thus, despite dnmt3a2 being a truncated form of dnmt3a2
and having similar activity in vitro, it is clear that they have very different functions in
vivo.
Like dnmt3a, dnmt3b preferentially methylates unmethylated DNA, and is
thought to function as a de novo methyltransferase (Okano et al., 1998). Dnmt3b is
highly expressed in ES cells, and Dnmt3b deficiency results in embryonic lethality
(Okano et al., 1999). In addition to general hypomethylation, dnmt3b -' - ES cells are
hypomethylated at their centromeric repeats, implying that dnmt3b plays a critical role in
this process (Okano et al., 1999; Xie et al., 1999). Dnmt3b appears to be partially
regulated through alternative splicing, with at least six splice variants having been
identified (Chen et al., 2002; Hansen et al., 1999; Okano et al., 1998). Of these six
alternative splice forms, only two encode for functional enzymatic domains, suggesting
that the other four lack methyltransferase activity. Somatically, dnmt3b is expressed at
low levels ubiquitously (Baylin et al., 2001), with a much higher level of expression in
the testis, ovary, liver, spleen and thymus (Chen et al., 2002). In humans, dnmt3b
mutations have been shown to be present in people with ICF syndrome (Hansen et al.,
1999). This disease is characterized by facial abnormalities and immunodeficiency.
Notably, individuals with this disease have hypomethylated centromeres and centromeric
instability (Ehrlich, 2003). Like dnmt3a, the role of dnmt3b in the regulation of somatic
methylation patterns is unknown.
There are two possible mechanisms for CpG demethylation. The first is through
passive demethylation.
In this instance, genes become demethylated during DNA
replication through an inhibition of maintenance methyltransferase.
Through this
process, the methylation level at a CpG could reduce by half with each cell division. The
other possible demethylation mechanism is active demethylation. In this instance, either
the cytosine is directly demethylated or a 5-methyl-cytosine is replaced with an
unmethylated cytosine independent of DNA replication. There is evidence that suggests
that widespread active demethylation of the paternal genome occurs in the mouse zygote,
without DNA replication (Santos et al., 2002).
There is also a report that active
demethylation occurs at the IL-2 promoter during T cell activation (Bruniquel and
38
Schwartz, 2003). There are no known active demethylation enzymes, and the chemical
demethylation of 5-methyl-cytosine is energetically unfavorable (Jeltsch, 2002), so the
mechanism behind active demethylation remains unknown.
There is strong evidence that DNA methylation plays an important role in helper
T cell polarization. TH1 cell lines, which express IFN-y but not IL-4, are hypomethylated
at the IFN-y locus, but hypermethylated at the IL-4 locus (Young et al., 1994). Likewise,
TH2 cell lines, which express IL-4 but not IFN-y, are hypomethylated at the IL-4 locus
but hypermethylated at the IFN-y locus (Young et al., 1994). Thus, the methylation
status of the helper T cell effector cytokine genes correlate well with gene expression in
fully polarized cell lines. Naive helper T cells, which do not transcribe IL-4, have been
shown to be hypermethylated at the IL-4 promoter (Lee et al., 2002). While naive helper
T cells also do not express IFN-y, there have been conflicting reports as to the
methylation of the promoter in these cells, with reports both stating that the promoter is
hypermethylated (Katamura et al., 1998; Melvin et al., 1995; White et al., 2002), and
hypomethylated (Winders et al., 2004). While the dynamics of methylation change
during helper T cell differentiation are unclear at the IFN-y promoter, the IL-4 promoter
appears to undergo a gradual demethylation that can be accounted for by passive
demethylation (Lee et al., 2002).
While demethylation is not required for IL-4
expression, there is a correlation between IL-4 promoter demethylation and a high level
of IL-4 expression (Lee et al., 2002).
In addition to the above correlations between effector cytokine gene methylation
status and transcription, knock-out mouse lines provide additional evidence that DNA
methylation plays a role in the regulation of cytokine expression during helper T cell
polarization. T cells from mice in which the maintenance methyltransferase, dnmtl, was
disrupted in thymocytes by Cre-mediated recombination had lower levels of methylation
in many genes, including IFN-y (Lee et al., 2001). T cells from these mice expressed a
higher level of both IFN-y and IL-4, consistent with the notion that methylation of the
cytokine genes acts to inhibit transcription (Lee et al., 2001).
Furthermore, mice
deficient in the methyl-CpG binding protein, MBD2, also show abnormal cytokine
expression (Hutchins et al., 2002). In the T cells from these mice, some cells polarized
under TH1 conditions expressed IL-4, in addition to IFN-y, while a significant fraction of
39
the cells polarized under TH2 conditions expressed IFN-y, in addition to IL-4 (Hutchins et
al., 2002). This suggests improper cytokine expression in polarized helper T cells is
partly repressed by the methyl-CpG binding protein MBD2.
Histone Modification and DNA Packaging
Eukaryotic genomic DNA is packaged in chromatin, the fundamental unit of
which is the nucleosome. This chromatin packaging is not uniform across the genome, as
some regions (euchromatin) are more loosely packaged, and thus more accessible to
transcription factors, while other regions (heterochromatin) are tightly packaged, and thus
inaccessible to transcription factors. In many parts of the genome the chromatin status is
dynamic, especially during cell differentiation (Richards and Elgin, 2002). Thus, altering
how DNA is packaged at a specific locus, and in doing so altering how accessible that
locus is to transcription factors, is a mechanism for the regulation of gene transcription.
There
are several factors
that contribute
to the differences
between
heterochromatin and euchromatin, including nucleosome density, histone composition,
and histone modifications (Richards and Elgin, 2002). Histones are subject to a wide
array of covalent modifications, including phosphorylation, methylation. acetylation, and
ubiquitination (Peterson and Laniel, 2004). It is thought that modifications to specific
amino acids on specific histones are able to impart a "histone code," resulting in the
recruitment of specific regulatory proteins to a locus that affect the overall gene activity
(Turner, 2002). While the effect of many of the modifications remains unclear, there are
certain modifications that are clearly correlated with gene transcription. Of particular
note, acetylation of histone H3 on lysine residues 9, 14, and 18, and on histone H4 on
lysine residue 5 are all associated with an accessible locus and active transcription
(Peterson and Laniel, 2004). Likewise, methylation of histone H3 lysine residue 4 is
associated with active transcription, while methylation of histone H3 on residues 9 or 27
is associated with inaccessible and transcriptionally silent genes (Zhang and Reinberg,
2001).
As mentioned above, histone modification and DNA methylation are interrelated.
It has been shown that methyl-CpG binding proteins are able to recruit histone
deacetylases (Jones et al., 1998; Nan et al., 1998) and histone methyltransferases (Fujita
40
et al., 2003; Fuks et al., 2003) to hypermethylated DNA. There is also evidence that, in
some species, the methylation lysine 9 of histone H3 can induce DNA methylation
(Lehnertz et al., 2003).
Thus, while there is a strong correlation between histone
modification and DNA methylation and it is clear that the regulation of these two events
is interrelated, many of the details of this relationship remain unknown.
One method of examining the chromatin status of a particular locus is by
determining site sensitivity to DNase I treatment. Areas with inaccessible chromatin
structure are resistant to DNase I degradation, while regions with accessible chromatin
structure are sensitive to DNase I degradation. This technique has proven particularly
informative about the chromatin changes that occur at the effector cytokine loci during
helper T cell differentiation. Of particular note, the appearance of DNase hypersensitive
(HS) sites located at the IL-4 and IFN-y promoters and at a 3' IL-4 enhancer correlate
well with the transcriptional activity of these genes, in that the IFN-y HS are present TH1
cells but not in naive helper T cells or TH2cells, while the IL-4 HS are present in TH2
cells, but not naive helper T cells or TH cells (Agarwal and Rao, 1998; Takemoto et al.,
1998). As the ectopic expression of T-bet in TH2 cells is able to induce the formation of
IFN-y HS (Mullen et al., 2002), while the ectopic expression of GATA-3 is able to induce
the formation of IL-4 HS (Lee et al., 2000), it is likely that these changes in chromatin
accessibility are mediated by the "master regulators" of helper T cell polarization.
It is also possible to directly measure histone modifications at a particular locus
through ChIP assays. This type of analysis revealed that histone H3 and H4 acetylation
correlates very strongly with transcription and HS formation at the effector cytokine loci.
In naive helper T cells both the IFN-y and IL-4 promoters are hypoacetylated, correlating
with the lack of gene transcription (Avni et al., 2002). Immediately following TCR
stimulation there is a nonspecific increase in the acetylation status at both the IFN-y and
IL-4 promoters (Avni et al., 2002). This increase in acetylation occurs independent of
polarization signals and correlates well with the observed polarization-independent
transcription of both IL-4 and IFN-y within 1 hour of TCR stimulation (Grogan et al.,
2001).
As THI1 differentiation
hyperacetlyatecl,
proceeds, the IFN-y promoter becomes further
while the IL-4 locus undergoes
deacetylation
and becomes
hypoacetylated (Avni et al., 2002). As TH2 differentiation proceeds, the IL-4 promoter
41
becomes hyperacetylated, while the IFN-y promoter becomes hypoacetylated (Avni et al.,
2002). These changes correlate well with the suppression of IL-4 transcription in TH2
cells and the suppression of IFN-y transcription in TH1 cells.
While there is a strong correlative evidence in support of epigenetic modification
playing an important role in the cytokine expression of helper T cells, very little is known
about the mechanisms through which these changes occur, and specifically, how histone
modification and DNA methyltransferase enzymes are targeted to particular loci by
specific polarization signals. Furthermore, while there is a strong correlation between
CpG methylation and gene silencing during this polarization process, how this silencing
is mediated is unknown.
In this thesis, I have examined the methylation changes that occur at the IFN-y
locus during helper T cell polarization. The site-specific nature of these changes was
analyzed by quantitatively measuring the methylation status of individual CpGs over
time. It was found that the kinetics of methylation change vary from CpG to CpG, with
the CpG located at the -53 position relative to the transcription start site becoming
methylated faster during TH2 polarization than the other CpGs of the proximal promoter.
The downstream effects of the methylation of the -53 CpG were looked at, and
methylation of the -53 position alone was found to block promoter binding by ATF2, cJun and CREB, and to inhibit promoter function.
Finally, the role of de novo
methyltransferases in these methylation changes was studied. Dnmt3a was found to play
a role in the de novo methylation of the promoter that occurs during TH2 polarization, and
clnmt3a2was found to play a role in the inhibition of promoter methylation during TH1
polarization.
42
References
Acuto, O., and Michel, F. (2003). CD28-mediated co-stimulation: a quantitative support
for TCR signalling. Nat Rev Immunol 3, 939-951.
Afkarian, M., Sedy, J. R., Yang, J., Jacobson, N. G., Cereb, N., Yang, S. Y., Murphy, T.
L., and Murphy, K. M. (2002). T-bet is a STAT 1-induced regulator of IL-12R expression
in naive CD4+ T cells. Nat Immunol 3, 549-557.
Agarwal, S., Avni, O., and Rao, A. (2000). Cell-type-restricted binding of the
transcription factor NFAT to a distal IL-4 enhancer in vivo. Immunity 12, 643-652.
Agarwal, S., and Rao, A. (1998). Modulation of chromatin structure regulates cytokine
gene expression during T cell differentiation. Immunity 9, 765-775.
Alegre, M. L., Frauwirth, K. A., and Thompson, C. B. (2001). T-cell regulation by CD28
and CTLA-4. Nat Rev Immunol 1, 220-228.
Alexopoulou, L., Holt, A. C., Medzhitov, R., and Flavell, R. A. (2001). Recognition of
double-stranded RNA and activation of NF-kappaB by Toll-like receptor 3. Nature 413,
732-738.
Alt, F. W., Oltz. E. M., Young, F., Gorman, J., Taccioli, G., and Chen, J. (1992). VDJ
recombination. Immunol Today 13, 306-314.
Ansel, K. M., Lee, D. U., and Rao, A. (2003). An epigenetic view of helper T cell
differentiation. Nat Immunol 4, 616-623.
Appleman, L. J., and Boussiotis, V. A. (2003). T cell anergy and costimulation. Immunol
Rev 192, 161-180.
Ardavin, C. (2003). Origin, precursors and differentiation of mouse dendritic cells. Nat
Rev Immunol 3, 582-590.
Avni, O., Lee, D., Macian, F., Szabo, S. J., Glimcher, L. H., and Rao, A. (2002). T(H)
cell differentiation is accompanied by dynamic changes in histone acetylation of cytokine
genes. Nat Immunol 3,643-651.
Banchereau, J., Briere, F., Caux, C., Davoust, J., Lebecque, S., Liu, Y. J., Pulendran, B.,
and Palucka, K. (2000). Immunobiology of dendritic cells. Annu Rev Immunol 18, 767811.
43
Barbulescu, K., Becker, C., Schlaak, J. F., Schmitt, E., Meyer zum Buschenfelde,
K. H.,
and Neurath, M. F. (1998). IL-12 and IL-18 differentially regulate the transcriptional
activity of the human IFN-gamma promoter in primary CD4+ T lymphocytes. J Immunol
160, 3642-3647.
Basson, M. A., and Zamoyska, R. (2000). The CD4/CD8 lineage decision: integration of
signalling pathways. Immunol Today 21, 509-514.
Baylin, S. B., Esteller, M., Rountree, M. R., Bachman, K. E., Schuebel, K., and Herman,
J. G. (2001). Aberrant patterns of DNA methylation, chromatin formation and gene
expression in cancer. Hum Mol Genet 10, 687-692.
Bestor, T., Laudano, A., Mattaliano, R., and Ingram, V. (1988). Cloning and sequencing
of a cDNA encoding DNA methyltransferase of mouse cells. The carboxyl-terminal
domain of the mammalian enzymes is related to bacterial restriction methyltransferases. J
Mol Biol 203,971-983.
Bestor, T. H. (1992). Activation of mammalian DNA methyltransferase by cleavage of a
Zn binding regulatory domain. Embo J 11, 2611-2617.
Biet, F., Locht, C., and Kremer, L. (2002). Immunoregulatory functions of interleukin 18
and its role in defense against bacterial pathogens. J Mol Med 80, 147-162.
Boehm, U., Klamp, T., Groot, M., and Howard, J. C. (1997). Cellular responses to
interferon-gamma. Annu Rev Immunol 15, 749-795.
Boring, L., Gosling, J., Chensue, S. W., Kunkel, S. L., Farese, R. V., Jr., Broxmeyer, H.
E., and Charo, I. F. (1997). Impaired monocyte migration and reduced type 1 (Thl)
cytokine responses in C-C chemokine receptor 2 knockout mice. J Clin Invest 100, 25522561.
Bour-Jordan, H., Grogan, J. L., Tang, Q., Auger, J. A., Locksley, R. M., and Bluestone, J.
A. (2003). CTLA-4 regulates the requirement for cytokine-induced signals in T(H)2
lineage commitment. Nat Immunol 4, 182-188.
Brod, S. A., Benjamin, D., and Hafler, D. A. (1991). Restricted T cell expression of IL-
2/IFN-gamma mRNA in human inflammatory disease. J Immunol 147, 810-815.
Bruniquel, D., and Schwartz, R. H. (2003). Selective, stable demethylation of the
interleukin-2 gene enhances transcription by an active process. Nat Immunol 4, 235-240.
Cantrell, D. (1996). T cell antigen receptor signal transduction pathways. Annu Rev
Immunol 14, 259-274.
44
Carpenter, G., and Ji, Q. (1999). Phospholipase C-gamma as a signal-transducing
element. Exp Cell Res 253, 15-24.
Chan, A. C., Dalton, M., Johnson, R., Kong, G. H., Wang, T., Thoma, R., and Kurosaki,
T. (1995). Activation of ZAP-70 kinase activity by phosphorylation of tyrosine 493 is
required for lymphocyte antigen receptor function. Embo J 14, 2499-2508.
Chatelain, R., Mauze, S., Varkila, K., and Coffman, R. L. (1999). Leishmania major
infection in interleukin-4 and interferon-gamma depleted mice. Parasite Immunol 21,
423-431.
Chen, Q., Ghilardi, N., Wang, H., Baker, T., Xie, M. H., Gurney, A., Grewal, I. S., and de
Sauvage, F. J. (2000). Development of Thl-type immune responses requires the type I
cytokine receptor TCCR. Nature 407, 916-920.
Chen, T., Ueda, Y., Xie, S., and Li, E. (2002). A novel Dnmt3a isoform produced from
an alternative promoter localizes to euchromatin and its expression correlates with active
de novo methylation. J Biol Chem 277, 38746-38754.
Chensue, S. W., Warmington, K. S., Ruth, J. H., Sanghi, P. S., Lincoln, P., and Kunkel,
S. L. (1996). Role of monocyte chemoattractant protein- (MCP- 1) in Th 1
(mycobacterial) and Th2 (schistosomal) antigen-induced granuloma formation:
relationship to local inflammation, Th cell expression, and IL-12 production. J Immunol
157, 4602-4608.
Cippitelli, M., Sica, A., Viggiano, V., Ye, J., Ghosh, P., Birrer, M. J., and Young, H. A.
(1995). Negative transcriptional regulation of the interferon-gamma promoter by
glucocorticoids and dominant negative mutants of c-Jun. J Biol Chem 270, 12548-12556.
Crabtree, G. R. 2001). Calcium, calcineurin, and the control of transcription. J Biol
Chem 276, 2313-2316.
Cross, S. H., Meehan, R. R., Nan, X., and Bird, A. (1997). A component of the
transcriptional repressor MeCP1 shares a motif with DNA methyltransferase and HRX
proteins. Nat Genet 16, 256-259.
de Carli, M., D'Elios, M. M., Mariotti, S., Marcocci, C., Pinchera, A., Ricci, M.,
Romagnani, S., and del Prete, G. (1993). Cytolytic T cells with Thl-like cytokine profile
predominate in retroorbital lymphocytic infiltrates of Graves' ophthalmopathy. J Clin
Endocrinol Metab 77, 1120-1124.
de Jong, E. C., Smits, H. H., and Kapsenberg, M. L. (2005). Dendritic cell-mediated T
cell polarization. Springer Semin Immunopathol 26, 289-307.
45
de Jong, E. C., Vieira, P. L., Kalinski, P., Schuitemaker, J. H., Tanaka, Y., Wierenga, E.
A., Yazdanbakhsh, M., and Kapsenberg, M. L. (2002). Microbial compounds selectively
induce Th cell-promoting or Th2 cell-promoting dendritic cells in vitro with diverse th
cell-polarizing signals. J Immunol 168, 1704-1709.
Dong, Z., Wang, X., and Evers, B. M. (2000). Site-specific DNA methylation contributes
to neurotensin/neuromedin N expression in colon cancers. Am J Physiol Gastrointest
Liver Physiol 279, G1 139-1147.
Dredge, K., Marriott, J. B., Todryk, S. M., and Dalgleish, A. G. (2002). Adjuvants and
the promotion of Thl-type cytokines in tumour immunotherapy. Cancer mmunol
Immunother 51.. 521-531.
Ebner, C., Szepfalusi, Z., Ferreira, F., Jilek, A., Valenta, R., Parronchi, P., Maggi, E.,
Romagnani, S., Scheiner, O., and Kraft, D. (1993). Identification of multiple T cell
epitopes on Bet v I, the major birch pollen allergen, using specific T cell clones and
overlapping peptides. J Immunol 150, 1047-1054.
Ehrlich, M. (2003). The ICF syndrome, a DNA methyltransferase 3B deficiency and
immunodeficiency disease. Clin Immunol 109, 17-28.
Finkelman, F. D., Shea-Donohue,
T., Morris, S. C., Gildea, L., Strait, R., Madden, K. B.,
Schopf, L., and Urban, J. F., Jr. (2004). Interleukin-4- and interleukin-13-mediated host
protection against intestinal nematode parasites. Immunol Rev 201, 139-155.
Flynn, S., Toellner, K. M., Raykundalia, C., Goodall, M., and Lane, P. (1998). CD4 T
cell cytokine differentiation: the B cell activation molecule, OX40 ligand, instructs CD4
T cells to express interleukin 4 and upregulates expression of the chemokine receptor,
Blr-l. J Exp Med 188, 297-304.
Fugmann, S. D. (2001). RAG1 and RAG2 in V(D)J recombination and transposition.
Immunol Res 23, 23-39.
Fujimoto, M., Kitazawa, R., Maeda, S., and Kitazawa, S. (2005). Methylation adjacent to
negatively regulating AP-1 site reactivates TrkA gene expression during cancer
progression. Oncogene 24, 5108-5118.
Fujita, N., Watanabe, S., Ichimura, T., Tsuruzoe, S., Shinkai, Y., Tachibana, M., Chiba,
T., and Nakao, M. (2003). Methyl-CpG binding domain (MBD1) interacts with the
Suv39h 1l-HPI heterochromatic complex for DNA methylation-based transcriptional
repression. J Biol Chem 278, 24132-24138.
46
Fuks, F., Hurd, P. J., Wolf, D., Nan, X., Bird, A. P., and Kouzarides, T. (2003). The
methyl-CpG-binding protein MeCP2 links DNA methylation to histone methylation. J
Biol Chem 278, 4035-4040.
Garcia, K. C., Teyton, L., and Wilson, I. A. (1999). Structural basis of T cell recognition.
Annu Rev Immunol 17, 369-397.
Genot, E., and Cantrell, D. A. (2000). Ras regulation and function in lymphocytes. Curr
Opin Immunol 12,289-294.
Godfrey, D. ., Kennedy, J., Mombaerts, P., Tonegawa, S., and Zlotnik, A. (1994). Onset
of TCR-beta gene rearrangement and role of TCR-beta expression during CD3-CD4CD8- thymocyte differentiation. J Immunol 152, 4783-4792.
Goronzy, J. J., and Weyand, C. M. (2005). Rheumatoid arthritis. Immunol Rev 204, 5573.
Gould, H. J., Sutton, B. J., Beavil, A. J., Beavil, R. L., McCloskey, N., Coker, H. A.,
Fear, D., and Smurthwaite, L. (2003). The biology of IGE and the basis of allergic
disease. Annu Rev Immunol 21, 579-628.
Grogan, J. L., and Locksley, R. M. (2002). T helper cell differentiation:
on again, off
again. Curr Opin Immunol 14, 366-372.
Grogan, J. L., Mohrs, M., Harmon, B., Lacy, D. A., Sedat, J. W., and Locksley, R. M.
(2001). Early transcription and silencing of cytokine genes underlie polarization of T
helper cell subsets. Immunity 14, 205-215.
Gromme, M., and Neefjes, J. (2002). Antigen degradation or presentation by MHC class I
molecules via classical and non-classical pathways. Mol Immunol 39, 181-202.
Hafler, D. A., Slavik, J. M., Anderson, D. E., O'Connor, K. C., De Jager, P., and Baecher-
Allan, C. (2005). Multiple sclerosis. Immunol Rev 204, 208-231.
Hai, T., and Curran, T. (1991). Cross-family dimerization of transcription factors Fos/Jun
and ATF/CREB alters DNA binding specificity. Proc Natl Acad Sci U S A 88, 37203724.
Hansen, R. S., Wijmenga, C., Luo, P., Stanek, A. M., Canfield, T. K., Weemaes, C. M.,
and Gartler, S. M. (1999). The DNMT3B DNA methyltransferase gene is mutated in the
ICF immunodeficiency syndrome. Proc Natl Acad Sci U S A 96, 14412-14417.
47
Hayashi, F., Smith, K. D., Ozinsky, A., Hawn, T. R., Yi, E. C., Goodlett. D. R., Eng, J.
K., Akira, S., Underhill, D. M., and Aderem, A. (2001). The innate immune response to
bacterial flagellin is mediated by Toll-like receptor 5. Nature 410, 1099-1103.
Heinzel, F. P., Sadick, M. D., and Locksley, R. M. (1988). Leishmania major: analysis of
lymphocyte and macrophage cellular phenotypes during infection of susceptible and
resistant mice. Exp Parasitol 65, 258-268.
Heinzel, F. P., Sadick, M. D., Mutha, S. S., and Locksley, R. M. (1991). Production of
interferon gamma, interleukin 2, interleukin 4, and interleukin 10 by CD4+ lymphocytes
in vivo during healing and progressive murine leishmaniasis. Proc Natl Acad Sci U S A
88, 7011-7015.
Hendrich, B., and Bird, A. (1998). Identification and characterization of a family of
mammalian methyl-CpG binding proteins. Mol Cell Biol 18, 6538-6547.
Hendriks, R. W., Nawijn, M. C., Engel, J. D., van Doorninck, H., Grosveld, F., and Karis,
A. (1999). Expression of the transcription factor GATA-3 is required for the development
of the earliest T cell progenitors and correlates with stages of cellular proliferation in the
thymus. Eur J Imnmunol29, 1912-1918.
Hess, J., Angel, P., and Schorpp-Kistner,
M. (2004). AP-1 subunits: quarrel and harmony
among siblings. J Cell Sci 117, 5965-5973.
Himmelrich, H., Parra-Lopez, C., Tacchini-Cottier,
F., Louis, J. A., and Launois, P.
(1998). The I L-4 rapidly produced in BALB/c mice after infection with Leishmania
major down-regulates IL-12 receptor beta 2-chain expression on CD4+ T cells resulting
in a state of unresponsiveness
to IL-12. J Immunol 161, 6156-6163.
Hirose, J., Kouro, T., Igarashi, H., Yokota, T., Sakaguchi, N., and Kincade, P. W. (2002).
A developing picture of lymphopoiesis in bone marrow. Immunol Rev 189, 28-40.
Ho, . C., Hodge, M. R., Rooney, J. W., and Glimcher, L. H. (1996). The proto-oncogene
c-maf is responsible for tissue-specific expression of interleukin-4. Cell 85, 973-983.
Hosken, N. A., Shibuya, K., Heath, A. W., Murphy, K. M., and O'Garra, A. (1995). The
effect of antigen dose on CD4+ T helper cell phenotype development in a T cell receptoralpha beta-transgenic model. J Exp Med 182, 1579-1584.
Hutchins, A. S., Mullen, A. C., Lee, H. W., Sykes, K. J., High, F. A., Hendrich, B. D.,
Bird, A. P., and Reiner, S. L. (2002). Gene silencing quantitatively controls the function
of a developmental trans-activator. Mol Cell 10, 81-91.
48
Hwang, E. S., Hong, J. H., and Glimcher, L. H. (2005a). IL-2 production in developing
Th cells is regulated by heterodimerization of RelA and T-bet and requires T-bet serine
residue 508. J Exp Med 202, 1289-1300.
Hwang, E. S., Szabo, S. J., Schwartzberg, P. L., and Glimcher, L. H. (2005b). T helper
cell fate specified by kinase-mediated interaction of T-bet with GATA-3. Science 307,
430-433.
Jacobson, N. G., Szabo, S. J., Weber-Nordt, R. M., Zhong, Z., Schreiber. R. D., Darnell,
J. E., Jr., and Murphy, K. M. (1995). Interleukin 12 signaling in T helper type 1 (Thl)
cells involves tyrosine phosphorylation of signal transducer and activator of transcription
(Stat)3 and Stat4. J Exp Med 181, 1755-1762.
Jain, J., Loh, C., and Rao, A. (1995). Transcriptional regulation of the IL-2 gene. Curr
Opin Immunol 7, 333-342.
.laneway, C. A., Jr., and Medzhitov, R. (2002). Innate immune recognition. Annu Rev
Immunol 20, 197-216.
Jankovic, D., Liu, Z., and Gause, W. C. (2001). Thl- and Th2-cell commitment during
infectious disease: asymmetry in divergent pathways. Trends Immunol 22, 450-457.
Jeltsch, A. (2002). Beyond Watson and Crick: DNA methylation and molecular
enzymology of DNA methyltransferases. Chembiochem 3, 274-293.
Jones, P. L., Veenstra, G. J., Wade, P. A., Vermaak, D., Kass, S. U., Landsberger, N.,
Strouboulis, J., and Wolffe, A. P. (1998). Methylated DNA and MeCP2 recruit histone
deacetylase to repress transcription. Nat Genet 19, 187-191.
Kagami, S., Nakajima, H., Suto, A., Hirose, K., Suzuki, K., Morita, S., Kato, I., Saito, Y.,
Kitamura, T., and Iwamoto, I. (2001). Stat5a regulates T helper cell differentiation by
several distinct mechanisms. Blood 97, 2358-2365.
Kalinski, P., Hilkens, C. M., Wierenga, E. A., and Kapsenberg, M. L. (1999). T-cell
priming by type--I and type-2 polarized dendritic cells: the concept of a third signal.
Immunol Today 20, 561-567.
Kaplan, M. H., Schindler, U., Smiley, S. T., and Grusby, M. J. (1996a). Stat6 is required
for mediating responses to IL-4 and for development of Th2 cells. Immunity 4, 313-319.
Kaplan, M. H., Sun, Y. L., Hoey, T., and Grusby, M. J. (1996b). Impaired IL-12
responses and enhanced development of Th2 cells in Stat4-deficient mice. Nature 382,
174-177.
49
Kapsenberg, M. L. (2003). Dendritic-cell control of pathogen-driven T-cell polarization.
Nat Rev Immunol 3, 984-993.
Katamura, K., Fukui, T., Kiyomasu, T., io, J., Tai, G., Ueno, H., Heike, T., Mayumi, M.,
and Furusho, K. (1998). IL-4 and prostaglandin E2 inhibit hypomethylation of the 5'
regulatory region of IFN-gamma gene during differentiation of naive CD4+ T cells. Mol
Immunol 35, 39-45.
Kato, T., and Nariuchi, H. (2000). Polarization of naive CD4+ T cells toward the Thl
subset by CTLA-4 costimulation. J Immunol 164, 3554-3562.
Kelly, M. A., Rayner, M. L., Mijovic, C. H., and Barnett, A. H. (2003). Molecular
aspects of type I diabetes. Mol Pathol 56, 1-10.
Kiani, A., Garcia-Cozar, F. J., Habermann, I., Laforsch, S., Aebischer, T., Ehninger, G.,
and Rao, A. (2001). Regulation of interferon-gamma gene expression by nuclear factor of
activated T cells. Blood 98, 1480-1488.
Kishikawa, H., Sun, J., Choi, A., Miaw, S. C., and Ho, I. C. (2001). The cell type-specific
expression of the murine IL-13 gene is regulated by GATA-3. J Immunol 167, 44144420.
Kloetzel, P. M. (2004). The proteasome and MHC class I antigen processing. Biochim
Biophys Acta 1695, 225-233.
Kong, G., Dalton, M., Wardenburg, J. B., Straus, D., Kurosaki, T., and Chan, A. C.
(1996). Distinct tyrosine phosphorylation sites in ZAP-70 mediate activation and
negative regulation of antigen receptor function. Mol Cell Biol 16, 5026-5035.
Kopf, M., Brombacher, F., Kohler, G., Kienzle, G., Widmann, K. H., Lefrang, K.,
Humborg, C., Ledermann, B., and Solbach, W. (1996). IL-4-deficient Balb/c mice resist
infection with Leishmania major. J Exp Med 184, 1127-1136.
Kupfer, A., and Singer, S. J. (1989). Cell biology of cytotoxic and helper T cell functions:
immunofluorescence microscopic studies of single cells and cell couples. Annu Rev
Immunol 7, 309-337.
Lee, D. U., Agarwal, S., and Rao, A. (2002). Th2 lineage commitment and efficient IL-4
production involves extended demethylation of the IL-4 gene. Immunity 16, 649-660.
Lee, H. J., Takemoto, N., Kurata, H., Kamogawa, Y., Miyatake, S., O'Garra, A., and
Arai, N. (2000). GATA-3 induces T helper cell type 2 (Th2) cytokine expression and
chromatin remodeling in committed Thl cells. J Exp Med 192, 105-115.
50
Lee, P. P., Fitzpatrick, D. R., Beard, C., Jessup, H. K., Lehar, S., Makar, K. W., PerezMelgosa, M., Sweetser, M. T., Schlissel, M. S., Nguyen, S., et al. (2001). A critical role
for Dnmtl and DNA methylation in T cell development, function, and survival.
Immunity 15, 763-774.
Lehnertz, B., Ueda, Y., Derijck, A. A., Braunschweig, U., Perez-Burgos, L., Kubicek, S.,
Chen, T., Li, E., Jenuwein, T., and Peters, A. H. (2003). Suv39h-mediated histone H3
lysine 9 methylation directs DNA methylation to major satellite repeats at pericentric
heterochromatin. Curr Biol 13, 1192-1200.
Leonhardt, H., Page, A. W., Weier, H. U., and Bestor, T. H. (1992). A targeting sequence
directs DNA methyltransferase to sites of DNA replication in mammalian nuclei. Cell 71,
865-873.
Li, E., Bestor, T. H., and Jaenisch, R. (1992). Targeted mutation of the DNA
methyltransferase gene results in embryonic lethality. Cell 69, 915-926.
Lighvani, A. A., Frucht, D. M., Jankovic, D., Yamane, H., Aliberti, J., Hissong, B. D.,
Nguyen, B. V., Gadina, M., Sher, A., Paul, W. E., and O'Shea, J. J. (2001). T-bet is
rapidly induced by interferon-gamma in lymphoid and myeloid cells. Proc Natl Acad Sci
U S A 98, 15137-15142.
Luk, J., Gross, P., and Thompson, W. W. (2001). Observations on mortality during the
1918 influenza pandemic. Clin Infect Dis 33, 1375-1378.
Macian, F. (2005). NFAT proteins: key regulators of T-cell development and function.
Nat Rev Immunol 5, 472-484.
Macian, F., Lopez-Rodriguez, C., and Rao, A. (2001). Partners in transcription: NFAT
and AP-1. Oncogene 20, 2476-2489.
Magram, J., Connaughton, S. E., Warrier, R. R., Carvajal, D. M., Wu, C. Y., Ferrante, J.,
Stewart, C., Sarmiento, U., Faherty, D. A., and Gately, M. K. (1996). IL-12-deficient
mice are defective in IFN gamma production and type 1 cytokine responses. Immunity 4,
471-481.
Martins, G. A., Hutchins, A. S., and Reiner, S. L. (2005). Transcriptional activators of
helper T cell fate are required for establishment but not maintenance of signature
cytokine expression. J Immunol 175, 5981-5985.
Melvin, A. J., McGurn, M. E., Bort, S. J., Gibson, C., and Lewis, D. B. (1995).
Hypomethylation of the interferon-gamma gene correlates with its expression by primary
T-lineage cells. Eur J Immunol 25, 426-430.
51
Misra, N., Murtaza, A., Walker, B., Narayan, N. P., Misra, R. S., Ramesh, V., Singh, S.,
Colston, M. J., and Nath, . (1995a). Cytokine profile of circulating T cells of leprosy
patients reflects both indiscriminate and polarized T-helper subsets: T-helper phenotype
is stable and uninfluenced by related antigens of Mycobacterium leprae. Immunology 86,
97-103.
Misra, N., Selvakumar, M., Singh, S., Bharadwaj, M., Ramesh, V., Misra, R. S., and
Nath, I. (1995b). Monocyte derived IL 10 and PGE2 are associated with the absence of
Th 1 cells and in vitro T cell suppression in lepromatous leprosy. Immunol Lett 48, 123128.
Mosmann, T. R., Cherwinski, H., Bond, M. W., Giedlin, M. A., and Coffman, R. L.
(1986). Two types of murine helper T cell clone. . Definition according to profiles of
lymphokine activities and secreted proteins. J Immunol 136, 2348-2357.
Mullen, A. C., High, F. A., Hutchins, A. S., Lee, H. W., Villarino, A. V., Livingston, D.
M., Kung, A. L., Cereb, N., Yao, T. P., Yang, S. Y., and Reiner, S. L. (2001). Role of T-
bet in commitment of TH cells before IL-12-dependent selection. Science 292, 19071910.
Mullen, A. C., Hutchins, A. S., High, F. A., Lee, H. W., Sykes, K. J., Chodosh, L. A., and
Reiner, S. L. (2002). Hlx is induced by and genetically interacts with T-bet to promote
heritable T(H)1 gene induction. Nat Immunol 3, 652-658.
Munder, M., Mallo, M., Eichmann, K., and Modolell, M. (1998). Murine macrophages
secrete interferon gamma upon combined stimulation with interleukin (IL)-12 and IL-18:
A novel pathway of autocrine macrophage activation. J Exp Med 187, 2103-2108.
Murphy, K. M., Ouyang, W., Farrar, J. D., Yang, J., Ranganath, S., Asnagli, H.,
Afkarian, M., and Murphy, T. L. (2000). Signaling and transcription in T helper
development. Annu Rev Immunol 18, 451-494.
Murphy, K. M., and Reiner, S. L. (2002). The lineage decisions of helper T cells. Nat
Rev Immunol 2, 933-944.
Nakahira, M., Ahn, H. J., Park, W. R., Gao, P., Tomura, M., Park, C. S., Hamaoka, T.,
Ohta, T., Kurimoto, M., and Fujiwara, H. (2002). Synergy of IL-12 and IL-18 for IFNgamma gene expression: IL- 12-induced STAT4 contributes to IFN-gamma promoter
activation by up-regulating the binding activity of IL-18-induced activator protein 1. J
Immunol 168, 1146-1153.
Nan, X., Ng, H. H., Johnson, C. A., Laherty, C. D., Turner, B. M., Eisenman, R. N., and
Bird, A. (1998). Transcriptional repression by the methyl-CpG-binding protein MeCP2
involves a histone deacetylase complex. Nature 393, 386-389.
52
Nelms, K., Keegan, A. D., Zamorano, J., Ryan, J. J., and Paul, W. E. (1999). The IL-4
receptor: signaling mechanisms and biologic functions. Annu Rev Immunol 17, 701-738.
Okano, M., Bell, D. W., Haber, D. A., and Li, E. (1999). DNA methyltransferases
Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development.
Cell 99, 247-257.
Okano, M., Xie, S., and Li, E. (1998). Cloning and characterization
of a family of novel
mammalian DNA (cytosine-5) methyltransferases. Nat Genet 19, 219-220.
Oppmann, B., Lesley, R., Blom, B., Timans, J. C., Xu, Y., Hunte, B., Vega, F., Yu, N.,
Wang, J., Singh, K., et al. (2000). Novel p19 protein engages IL-12p40 to form a
cytokine, IL-23, with biological activities similar as well as distinct from IL-12.
Immunity 13, 715-725.
Ouyang, W., Lohning, M., Gao, Z., Assenmacher, M., Ranganath, S., Radbruch, A., and
Murphy, K. M. (2000). Stat6-independent GATA-3 autoactivation directs IL-4independent Th2 development and commitment. Immunity 12, 27-37.
Ouyang, W., Ranganath, S. H., Weindel, K., Bhattacharya, D., Murphy, T. L., Sha, W.
C., and Murphy, K. M. (1998). Inhibition of Thl development mediated by GATA-3
through an IL-4-independent mechanism. Immunity 9, 745-755.
Pai, S. Y., Truitt, M. L., and Ho, I. C. (2004). GATA-3 deficiency abrogates the
development and maintenance of T helper type 2 cells. Proc Natl Acad Sci U S A 101,
1993-1998.
Palmer, E. (2003). Negative selection--clearing out the bad apples from the T-cell
repertoire. Nat Rev Immunol 3, 383-391.
Pandolfi, P. P., Roth, M. E., Karis, A., Leonard, M. W., Dzierzak, E., Grosveld, F. G.,
Engel, J. D., and Lindenbaum, M. H. (1995). Targeted disruption of the GATA3 gene
causes severe abnormalities in the nervous system and in fetal liver haematopoiesis. Nat
Genet 11, 40-44.
Pearce, E. J., Caspar, P., Grzych, J. M., Lewis, F. A., and Sher, A. (1991).
D)ownregulation of Th 1 cytokine production accompanies induction of Th2 responses by
a parasitic helminth, Schistosoma mansoni. J Exp Med 173, 159-166.
Penix, L., Weaver, W. M., Pang, Y., Young, H. A., and Wilson, C. B. (1993). Two
essential regulatory elements in the human interferon gamma promoter confer activation
specific expression in T cells. J Exp Med 178, 1483-1496.
53
Penix, L. A., Sweetser, M. T., Weaver, W. M., Hoeffler, J. P., Kerppola, T. K., and
Wilson, C. B. (1996). The proximal regulatory element of the interferon-gamma
promoter mediates selective expression in T cells. J Biol Chem 271, 31964-31972.
Pestka, S., Kotenko, S. V., Muthukumaran,
G., Izotova, L. S., Cook, J. R., and Garotta,
G. (1997). The interferon gamma (IFN-gamma) receptor: a paradigm for the multichain
cytokine receptor. Cytokine Growth Factor Rev 8, 189-206.
Peterson, C. L., and Laniel, M. A. (2004). Histones and histone modifications. Curr Biol
14, R546-551.
Pleiman, C. M., D'Ambrosio, D., and Cambier, J. C. (1994). The B-cell antigen receptor
complex: structure and signal transduction. Immunol Today 15, 393-399.
Rao, A., Luo, C., and Hogan, P. G. (1997). Transcription factors of the NFAT family:
regulation and function. Annu Rev Immunol 15, 707-747.
Rengarajan, J., Tang, B., and Glimcher, L. H. (2002). NFATc2 and NFATc3 regulate
T(H)2 differentiation and modulate TCR-responsiveness of naive T(H)cells. Nat
Immunol 3, 48-54.
Richards, E. J., and Elgin, S. C. (2002). Epigenetic codes for heterochromatin formation
and silencing: rounding up the usual suspects. Cell 108, 489-500.
Romagnani, S. (1994). Lymphokine production by human T cells in disease states. Annu
Rev Immunol 12, 227-257.
Rulifson, I. C., Sperling, A. I., Fields, P. E., Fitch, F. W., and Bluestone, J. A. (1997).
CD28 costimulation promotes the production of Th2 cytokines. J Immunol 158, 658-665.
Rusnak, F., and Mertz, P. (2000). Calcineurin: form and function. Physiol Rev 80, 14831521.
Sadick, M. D., Locksley, R. M., Tubbs, C., and Raff, H. V. (1986). Murine cutaneous
leishmaniasis: resistance correlates with the capacity to generate interferon-gamma in
response to Leishmania antigens in vitro. J Immunol 136, 655-661.
Saint-Ruf, C., Ungewiss, K., Groettrup, M., Bruno, L., Fehling, H. J., and von Boehmer,
H. (1994). Analysis and expression of a cloned pre-T cell receptor gene. Science 266,
1208-1212.
Samelson, L. E. (2002). Signal transduction mediated by the T cell antigen receptor: the
role of adapter proteins. Annu Rev Immunol 20, 371-394.
54
Santos, F., Hendrich, B., Reik, W., and Dean, W. (2002). Dynamic reprogramming of
DNA methylation in the early mouse embryo. Dev Biol 241, 172-182.
Scott, P., Pearce, E., Cheever, A. W., Coffman, R. L., and Sher, A. (1989). Role of
cytokines and CD4+ T-cell subsets in the regulation of parasite immunity and disease.
Immunol Rev 112, 161-182.
Seder, R. A., and Ahmed, R. (2003). Similarities and differences in CD4+ and CD8+
effector and memory T cell generation. Nat Immunol 4, 835-842.
Shi, Y., Seto, E., Chang, L. S., and Shenk, T. (1991). Transcriptional
repression by YY 1,
a human GLI-Kruppel-related protein, and relief of repression by adenovirus E1A
protein. Cell 67, 377-388.
Shortman, K., and Wu, L. (1996). Early T lymphocyte progenitors. Annu Rev Immunol
14, 29-47.
Siegfried, Z., and Cedar, H. (1997). DNA methylation: a molecular lock. Curr Biol 7,
R305-307.
Sleckman, B. P., Bassing, C. H., Bardon, C. G., Okada, A., Khor, B., Bories, J. C.,
Monroe, R., and Alt, F. W. (1998). Accessibility control of variable region gene assembly
during T-cell development. Immunol Rev 165, 121-130.
Soutto, M., Zhang, F., Enerson, B., Tong, Y., Boothby, M., and Aune, T. M. (2002). A
minimal IFN-gamma promoter confers Thl selective expression. J Immunol 169, 42054212.
Spellberg, B., and Edwards, J. E., Jr. (2001). Type 1/Type 2 immunity in infectious
diseases. Clin Infect Dis 32, 76-102.
Stein, R., Gruenbaum, Y., Pollack, Y., Razin, A., and Cedar, H. (1982). Clonal
inheritance of the pattern of DNA methylation in mouse cells. Proc Natl Acad Sci U S A
79, 61-65.
Straus, D. B., and Weiss, A. (1992). Genetic evidence for the involvement of the lck
tyrosine kinase in signal transduction through the T cell antigen receptor. Cell 70, 585593.
Szabo, S. J., Dighe, A. S., Gubler, U., and Murphy, K. M. (1997). Regulation of the
interleukin (IL)-12R beta 2 subunit expression in developing T helper 1 (Thl) and Th2
cells. J Exp Med 185, 817-824.
55
Szabo, S. J., Kim, S. T., Costa, G. L., Zhang, X., Fathman, C. G., and Glimcher, L. H.
(2000). A novel transcription factor, T-bet, directs Th 1 lineage commitment. Cell 100,
655-669.
Szabo, S. J., Sullivan, B. M., Stemmann, C., Satoskar, A. R., Sleckman, B. P., and
Glimcher, L. H. (2002). Distinct effects of T-bet in TH lineage commitment and IFNgamma production in CD4 and CD8 T cells. Science 295, 338-342.
Takeda, A., Hamano, S., Yamanaka, A., Hanada, T., Ishibashi, T., Mak, T. W.,
Yoshimura, A., and Yoshida, H. (2003). Cutting edge: role of IL-27/WSX-1 signaling for
induction of T-bet through activation of STAT1 during initial Thl commitment. J
Immunol 170, 4886-4890.
Takeda, K., Tanaka, T., Shi, W., Matsumoto, M., Minami, M., Kashiwamura, S.,
Nakanishi, K., Yoshida, N., Kishimoto, T., and Akira, S. (1996). Essential role of Stat6 in
IL-4 signalling. Nature 380, 627-630.
Takemoto, N., Kamogawa, Y., Jun Lee, H., Kurata, H., Arai, K. I., O'Garra, A., Arai, N.,
and Miyatake, S. (2000). Cutting edge: chromatin remodeling at the IL-4/IL-13 intergenic
regulatory region for Th2-specific cytokine gene cluster. J Immunol 165, 6687-6691.
Takemoto, N., Koyano-Nakagawa,
N., Yokota, T., Arai, N., Miyatake, S., and Arai, K.
(1998). Th2-specific DNase I-hypersensitive sites in the murine IL-13 and IL-4 intergenic
region. Int Immunol 10, 1981-1985.
Tan, S. L., and Parker, P. J. (2003). Emerging and diverse roles of protein kinase C in
immune cell signalling. Biochem J 376, 545-552.
Teixeira, L. K., Fonseca, B. P., Vieira-de-Abreu, A., Barboza, B. A., Robbs, B. K.,
Bozza, P. T., and Viola, J. P. (2005). IFN-gamma production by CD8+ T cells depends
on NFAT 1 transcription factor and regulates Th differentiation. J Immunol 175, 59315939.
Thomas, M. J., and Seto, E. (1999). Unlocking the mechanisms of transcription factor
YY 1: are chromatin modifying enzymes the key? Gene 236, 197-208.
Tivol, E. A., Borriello, F., Schweitzer, A. N., Lynch, W. P., Bluestone, J. A., and Sharpe,
A. H. (1995). Loss of CTLA-4 leads to massive lymphoproliferation and fatal multiorgan
tissue destruction, revealing a critical negative regulatory role of CTLA-4. Immunity 3,
541-547.
56
Tong, Y., Aune, T., and Boothby, M. (2005). T-bet antagonizes mSin3a recruitment and
transactivates a fully methylated IFN-gamma promoter via a conserved T-box half-site.
Proc Natl Acad Sci U S A 102, 2034-2039.
Trinchieri, G. (2003). Interleukin-12 and the regulation of innate resistance and adaptive
immunity. Nat Rev Immunol 3, 133-146.
Trinchieri, G., Pflanz, S., and Kastelein, R. A. (2003). The IL-12 family of heterodimeric
cytokines: new players in the regulation of T cell responses. Immunity 19, 641-644.
Turner, B. M. (2002). Cellular memory and the histone code. Cell 111, 285-291.
Usui, T., Nishikomori, R., Kitani, A., and Strober, W. (2003). GATA-3 suppresses Thl
development by downregulation of Stat4 and not through effects on IL-12Rbeta2 chain or
T-bet. Immunity 18, 415-428.
Villadangos, J. A., and Ploegh, H. L. (2000). Proteolysis in MHC class II antigen
presentation: who's in charge? Immunity 12, 233-239.
von Boehmer, FI. (2005). Unique features of the pre-T-cell receptor alpha-chain: not just
a surrogate. Nat Rev Immunol 5, 571-577.
von Boehmer, H., Aifantis, I., Gounari, F., Azogui, O., Haughn, L., Apostolou, I.,
Jaeckel, E., Grassi, F., and Klein, L. (2003). Thymic selection revisited: how essential is
it? Immunol Rev 191, 62-78.
Wange, R. L., Guitian, R., Isakov, N., Watts, J. D., Aebersold, R., and Samelson, L. E.
(1995). Activating and inhibitory mutations in adjacent tyrosines in the kinase domain of
ZAP-70. J Biol Chem 270, 18730-18733.
Wange, R. L., and Samelson, L. E. (1996). Complex complexes: signaling at the TCR.
Immunity 5, 197-205.
Waterhouse, P., Penninger, J. M., Timms, E., Wakeham, A., Shahinian, A., Lee, K. P.,
Thompson, C. B., Griesser, H., and Mak, T. W. (1995). Lymphoproliferative disorders
with early lethality in mice deficient in Ctla-4. Science 270, 985-988.
Watford, W. T., Hissong, B. D., Bream, J. H., Kanno, Y., Muul, L., and O'Shea, J. J.
(2004). Signaling by IL-12 and IL-23 and the immunoregulatory roles of STAT4.
Immunol Rev 202, 139-156.
Watts, C. (2004). The exogenous pathway for antigen presentation on major
histocompatibility complex class II and CD1 molecules. Nat Immunol 5, 685-692.
57
Weisenberger,
D. J., Velicescu, M., Preciado-Lopez,
M. A., Gonzales, F. A., Tsai, Y. C.,
Liang, G., and Jones, P. A. (2002). Identification and characterization of alternatively
spliced variants of DNA methyltransferase 3a in mammalian cells. Gene 298, 91-99.
Wenner, C. A., Guler, M. L., Macatonia, S. E., O'Garra, A., and Murphy, K. M. (1996).
Roles of IFN-gamma and IFN-alpha in IL-12-induced T helper cell-I development. J
Immunol 156, 1442-1447.
White, G. P., Watt, P. M., Holt, B. J., and Holt, P. G. (2002). Differential patterns of
methylation of the IFN-gamma promoter at CpG and non-CpG sites underlie differences
in IFN-gamma gene expression between human neonatal and adult CD45RO- T cells. J
Immunol 168, 2820-2827.
Wierenga, E. A., Snoek, M., Bos, J. D., Jansen, H. M., and Kapsenberg, M. L. (1990).
Comparison of diversity and function of house dust mite-specific T lymphocyte clones
from atopic and non-atopic donors. Eur J Immunol 20, 1519-1526.
Winders, B. R., Schwartz, R. H., and Bruniquel, D. (2004). A distinct region of the
murine IFN-gamma promoter is hypomethylated from early T cell development through
mature naive and Thl cell differentiation, but is hypermethylated in Th2 cells. J Immunol
173, 7377-7384.
Wu, L., Antica, M., Johnson, G. R., Scollay, R., and Shortman, K. (1991).
Developmental potential of the earliest precursor cells from the adult mouse thymus. J
Exp Med 174, 1617-1627.
Wurster, A. L., Tanaka, T., and Grusby, M. J. (2000). The biology of Stat4 and Stat6.
Oncogene 19, 2577-2584.
Xie, S., Wang, Z., Okano, M., Nogami, M., Li, Y., He, W. W., Okumura, K., and Li, E.
(1999). Cloning, expression and chromosome locations of the human DNMT3 gene
family. Gene 236, 87-95.
Xu, X., Sun, Y. L., and Hoey, T. (1996). Cooperative DNA binding and sequence-
selective recognition conferred by the STAT amino-terminal domain. Science 273, 794797.
Yang, W. M., Inouye, C., Zeng, Y., Bearss, D., and Seto, E. (1996). Transcriptional
repression by YY 1 is mediated by interaction with a mammalian homolog of the yeast
global regulator RPD3. Proc Natl Acad Sci U S A 93, 12845-12850.
58
Yang, W. M., Yao, Y. L., Sun, J. M., Davie, J. R., and Seto, E. (1997). Isolation and
characterization of cDNAs corresponding to an additional member of the human histone
deacetylase gene family. J Biol Chem 272, 28001-28007.
Yazdanbakhsh,
M., Kremsner, P. G., and van Ree, R. (2002). Allergy, parasites, and the
hygiene hypothesis. Science 296, 490-494.
Ye, J., Cippitelli, M., Dorman, L., Ortaldo, J. R., and Young, H. A. (1996). The nuclear
factor YY 1 suppresses the human gamma interferon promoter through two mechanisms:
inhibition of AP1 binding and activation of a silencer element. Mol Cell Biol 16, 47444753.
Yoshida, H., Hamano, S., Senaldi, G., Covey, T., Faggioni, R., Mu, S., Xia, M.,
Wakeham, A. C., Nishina, H., Potter, J., et al. (2001). WSX-1 is required for the
initiation of Th 1responses and resistance to L. major infection. Immunity 15, 569-578.
Yoshimura, A., Lien, E., Ingalls, R. R., Tuomanen, E., Dziarski, R., and Golenbock, D.
(1999). Cutting edge: recognition of Gram-positive bacterial cell wall components by the
innate immune system occurs via Toll-like receptor 2. J Immunol 163, 1-5.
Young, H. A., Ghosh, P., Ye, J., Lederer, J., Lichtman, A., Gerard, J. R., Penix, L.,
Wilson, C. B., Melvin, A. J., McGurn, M. E., and et al. (1994). Differentiation of the T
helper phenotypes by analysis of the methylation state of the IFN-gamma gene. J
Immunol 153, 3603-3610.
Zhang, D. H., Cohn, L., Ray, P., Bottomly, K., and Ray, A. (1997). Transcription factor
GATA-3 is differentially expressed in murine Thl and Th2 cells and controls Th2specific expression of the interleukin-5 gene. J Biol Chem 272, 21597-21603.
Zhang, F., Wang, D. Z., Boothby, M., Penix, L., Flavell, R. A., and Aune, T. M. (1998a).
Regulation of the activity of IFN-gamma promoter elements during Th cell
differentiation. J Immunol 161, 6105-6112.
Zhang, W., Sloan-Lancaster,
J., Kitchen, J., Trible, R. P., and Samelson, L. E. (1998b).
LAT: the ZAP-70 tyrosine kinase substrate that links T cell receptor to cellular activation.
Cell 92, 83-92.
Zhang, Y., and Reinberg, D. (2001). Transcription regulation by histone methylation:
interplay between different covalent modifications of the core histone tails. Genes Dev
15, 2343-2360.
Zhu, J., Cote-Sierra, J., Guo, L., and Paul, W. E. (2003). StatS activation plays a critical
role in Th2 differentiation. Immunity 19, 739-748.
59
Zhu, J., Min, B., Hu-Li, J., Watson, C. J., Grinberg, A., Wang, Q., Killeen, N., Urban, J.
F., Jr., Guo, L., and Paul, W. E. (2004). Conditional deletion of Gata3 shows its essential
function in T(H)l-T(H)2 responses. Nat Immunol 5, 1157-1165.
60
Chapter 2:
Site-Specific Methylation Suppresses IFN-y Promoter
Activity by Directly Inhibiting Factor Binding
Brendan Jones and Jianzhu Chen
61
Summary
When they become activated, naive helper T cells are able to polarize into either TH
cells or TH2 cells. Development of naive CD4+ T cells into TH1 cells is characterized by
the expression of IFN-y and the silencing of IL-4, while development into TH2 cells is
characterized by expression of IL-4 and silencing of IFN-y. Here we show that the IFN-y
proximal promoter undergoes progressive de novo methylation during TH2 polarization.
Notably, methylation at different CpG sites occurs at different rates, with methylation
occurring fastest at the CpG located at the -53 position relative to the transcription start
site. Methylation at this position inhibits c-Jun, ATF2 and CREB binding in vitro.
Consistently, the same factors bind to the unmethylated promoter in a TH1 cell line, but
not the methylated promoter in a TH2 cell line. Furthermore, methylation of the proximal
promoter at the -53 position alone is sufficient to inhibit promoter activity in transient
transfection assays. Thus, the rapid methylation of the -53 CpG at the onset of TH2
polarization helps to prevent IFN-y transcription by directly inhibiting transcription factor
binding prior to the extensive methylation of the IFN-y promoter.
62
Introduction
One of the most important aspects of the adaptive immune system is the ability to
adjust its response in order to more effectively combat different forms of pathogenic
infection.
An important mechanism that contributes to this adaptive process is the
polarization of CD4+ helper T cells. When a naive helper T cell becomes activated, it can
differentiate into at least two phenotypes, T helper type 1 (TH1),or T helper type 2 (TH2)
(Jankovic et al., 2001; Mosmann et al., 1986; Murphy and Reiner, 2002). T 1 cells are
generated in response to intracellular pathogens, such as virus or intracellular bacteria
(Romagnani, 1994; Spellberg and Edwards, 2001), and secrete interferon-y (IFN-y),
tumor necrosis factor a (TNFa) and TNF3 (Grogan and Locksley, 2002; Murphy et al.,
2000). TH2 cells are generated in response to helminths (Pearce et al., 1991; Scott et al.,
1989), and secrete a different set of cytokines, including interleukin-4 (IL-4), IL-5, and
IL-13 (Romagnani, 1994; Spellberg and Edwards, 2001).
Improper helper T cell
polarization has been shown result in an ineffective immune response, allergies, and
autoimmune disease (Chitnis and Khoury, 2003; Romagnani, 1994; Spellberg and
Edwards, 2001).
The molecular and genetic interactions involved in the regulation of this
polarization process have been extensively studied, and have been described in detail in
the previous chapter. One of the major factors that determines naive helper T cell fate
upon antigenic stimulation is the induction of transcription factors T-bet and GATA3
(Lee et al., 2000; Szabo et al., 2000; Zhang et al., 1997). Expression of T-bet, which is
downstream of the IFN-y receptor (Lighvani et al., 2001) and Statl (Mullen et al., 2001),
is sufficient to induce TH1 polarization of naive CD4+ T cells, and ectopic expression in
fully differentiated TH2 cells is sufficient to induce IFN-y expression upon PMA and
ionomycin stimulation (Szabo et al., 2000). Furthermore, mice that are deficient in T cell
T-bet expression have impaired T 1 polarization (Szabo et al., 2002). Expression of
GATA3, which has been shown to be downstream of the IL-4 receptor and Stat6 (Zhang
et al., 1997), is sufficient to induce a TH2 phenotype in CD4 + T cells (Lee et al., 2000),
and mice that are deficient in T cell GATA3 expression have an impaired, though not
completely depleted, TH2 response (Pai et al., 2004; Zhu et al., 2004). Furthermore, T-bet
63
suppresses TH2 polarization, at least partly through a direct binding and inhibition of
GATA3 (Hwang et al., 2005), while GATA3 represses TH polarization by inhibiting
Stat4, (Usui et al., 2003) and suppressing IL-12 receptor beta expression (Lee et al.,
2000). Because of this, once a sufficient level of T-bet or GATA3 is induced, a T cell
will polarize to one of the two subsets.
While T-bet and GATA3 play a major roll in the differentiation of helper T cells,
the upregulation
of these transcription
factors are not necessary for the initial
transcription of IFN-y and IL-4. It has been demonstrated that, under both THI and TH2
polarizing conditions, both IFN-y and IL-4 transcription is initiated within one hour of
naive helper T cell activation (Grogan et al., 2001). In contrast, T-bet and GATA3
mRNA levels are not significantly increased until between 24 and 72 hours postactivation (Grogan et al., 2001). These observations indicated that IFN-y and IL-4 can be
expressed without upregulation of T-bet or GATA3 respectively. Furthermore, factors
sufficient for transcription of IFN-y and IL-4 preexist in naive helper T cells, and, upon
activation, conditions temporarily become conducive to the expression of both cytokines.
A sustained expression of either cytokine following CD4+ T cell activation would likely
interfere with helper T cell polarization because IFN-y promotes TH polarization but
inhibits TH2 polarization (Afkarian et al., 2002; Seder et al., 1992), and 11-4promotes TH2
polarization
but inhibits
TH
polarization
(Hsieh et al., 1992; Seder et al., 1992). Thus,
proper helper T cell polarization probably requires a mechanism to suppress the
inappropriate cytokine gene expression. Indeed, as CD4+ T cell polarization progresses,
IFN-y transcription is repressed in cells polarized under TH2 conditions, while
transcription of IL-4 is repressed in cells polarized under THI conditions (Grogan et al.,
2001).
As was described in the previous chapter, a mechanism that has been implicated
in long-term suppression of IFN-y expression is DNA methylation (Yano et al., 2003;
Young et al., 1994) and chromatin remodeling
(Agarwal and Rao, 1998; Ansel et al.,
2003). CpG methylation in the promoter region often correlates with gene expression in
that transcribed genes are usually hypomethylated whereas silent genes are usually
hypermethylated (Jeltsch, 2002; Richards and Elgin, 2002). Consistently, the IFN-y locus
is hypomethylated
in IFN-y expressing THI cell lines but hypermethylated
in TH2 cell
64
lines, which do not express IFN-y (Young et al., 1994). In naive CD4+ T cells, from
which the THI and TH2 populations arise, there are conflicting reports as to the
methylation status of IFN-y promoter, as it has been reported as both hypermethylated
(Katamura
et al.,
1998; Melvin
et al.,
1995; White
et al.,
2002)
and
hypomethylated(Winders et al., 2004).
It has been suggested that hypermethylation could induce chromatin remodeling
through a variety of mechanisms, including the recruitment of methylated CpG binding
proteins and histone deacetylases (Jones et al., 1998; Nan et al., 1998). As histone
hyperacetylation is associated with an open chromatin confirmation, while histone
hypoacetylation is associated with a closed chromatin confirmation (Jenuwein and Allis,
2001), histone cleacetylases provide a mechanism for the induction of chromatin mediated
silencing.
It has been shown that the IFN-y locus is hypoacetylated
in nafve helper T
cells (Avni et al., 2002), correlating with a lack of IFN-y expression in these cells. Upon
T cell activation, the IFN-y locus rapidly becomes acetylated independent of polarizing
conditions (Avni et al., 2002), which correlates with the rapid T-bet independent IFN-y
transcription that occurs following activation (Grogan et al., 2001).
During TH1
polarization, the locus remains hypoacetylated, while during TH2 polarization the locus
becomes hyperacetylated (Avni et al., 2002). This correlates both with the IFN-y
transcription and the promoter methylation state in these populations. Thus, promoter
methylation provides a possible mechanism for the chromatin-mediated silencing of the
IFN-y locus that has been observed in long-term polarized TH2 cells.
Because establishment of chromatin-mediated transcription suppression is usually
a slow process, the question that arises is whether methylation at the IFN-y promoter
directly and immediately inhibits IFN-y transcription by inhibiting transcription factor
binding during the course of TH2 polarization. Among the transcription factors where
this type of regulation is likely are the AP1 and CREB/ATF families. There are at least
two possible binding sites for AP1 and CREB/ATF family members in the proximal IFNy promoter, one located at -90 to -78 bp (distal) and the other at -70 to -44 bp (proximal)
relative to the transcription start site (Figure 1) (Aune et al., 1997; Penix et al., 1996;
Zhang et al., 1998), with both sites capable of driving some level of reporter gene
expression in cells that express IFN-y (Aune et al., 1997; Penix et al., 1993).
65
-205 -190 -170
-52
-45
-34
+16
+96'
A.
B.
IFN-y proximal promoter sequence alignment
Mouse
Rat
Human
Chimp
Dog
-205
-190
-170
TTCAGAGAATCCCACAAGAATGGCACAGGTGGGCACAGCGGGGCTGTCTCATCGTCAGAGAGCCCAAGGAGTCGAAAGGA
TTCAGAGAATCCCACAAGAATGGCACAGGTGGGCACGGTGGGACCGTCTCATCGTCAGAGAGCCCAAGGAGTCGAAAGGA
TTCAAAGAATCCCACCAGAATGGCACAGGTGGGCATAATGGGTCTGTCTCATCGTCAAAGGACCCAAGGAGTCTAAAGGA
TTCAAAGAATCCCACCAGAATGGCACAGGTGGGCATAATGGGTCTGTCTCATCGTCAAAGGACCCAAGGAGTCTAAAGGA
TTCAAAGGATCCCACAAGAATGGCACAGAGGGGCATAATGGGTCTGTCTCATCGTCAAAAGACCCAAGGAGTTGAAAGGA
Mouse
Rat
Human
Chimp
Dog
AACTCTAAC----ATGCC-ACA------AAA-CCATAGCTGTAATGCAAAGTAACTTAGCTCCCCCCACCTATCTGTCAC
AACTCTAACTACAACACCGACAGGCCACAAA-CCATAGCTATAATGCAAAGTAACTAAGCTCCCGCCACCTATCTTTCAC
AACTCTAACTACAACACCCAAATGCCACAAAACCTTAGTTAT----TAATACAAACTATCATCC-CTGCCTATCTGTCAC
AACTCTAACTACAACACCCAAATGCCACAAAACCTTAGTTAT----TAATACAAACTAGCATCC-CTGCCTATCTGTCAC
AACTCTAACTACAACACCAAAATGCCACAAAACCATAGTTAT----TAATACAAACTAGCATCT-CTGCCTATCTGTCAC
Mouse
Rat
Human
Chimp
Dog
-53
45
-34
CATCTTAAAAAAAAAAAAAAACCAAAAAAAAACTTG
AAAATACGTAAT CCGAGGAGCCTTCGATCAGGTATAAA-AC
CATCTTAAC ----------TTAAAAAAAAACCTGW
AAAATACGTAAT¶CCAAGAAGCCTTCGGTCATGTATAAA-AC
CATCTCA--------------TCTTAAAAAA-CTTGWAAAATACGTAATCTCAGGAGACTTCAATTAGGTATAAATAC
CATCTCA--------------TCTTAAAAAA-CTTGWAAAATACGTAATCTCAGGAGACTTCAATTAGGTATAAATAC
CATCTCA--------------GCTAAAAAAAACTTG1TAAAATACGTAA
CT--GAGGACTTCAATTAGGTATAAATAC
CREB/ATF AP1
Mouse
Rat
Human
Chimp
Dog
Mouse
Rat
Human
Chimp
Dog
+16
I
TGGAAGCCAGAG-AGGTGCAGGCTATAGCT---GCCATC----GGCTGACC--TAGAGAAGACACATCAGCTGA-TCCTT
TGGAAGCAAGAG-AGGTGCAGCGTATAGCT---GCCATC---- GGCTGATC--TAGAGAAGACACATCAGCTGA-TTCTT
CAGCAGCCAGAGGAGGTGCAGCACATTGTTCTGATCATCTGAAGATCAGCTATTAGAAGAGAAAGATCAGTTAAGTCCTT
CAGCAGCCAGAGGAGGTGCAGCACATTGTTCTGATCATCTGAAGATCAGCTATTAGAAGAGAAAGATCAGTTAAGTCCTT
CAGCAGCAGAAGGAGATGCAGTACATTTTTCTGATTGTCTACAGGTTGGCTATTAGAAAAGAAAGATCAGCTGAGTCCTT
+96
TGGACCCT-CTGACTTGAGACAGAAGTTCTGGGCTTCTCCTCCTGCGG
CGGACTCT-CTGACTTAATACAGGAGTTCTGGGCTTTCCCTGCCTTGG
TGGACCTGATCAGCTTGATACAAGAACTACTGATTTCAACTTCTTTGG
TGGACCTGATCAGCTTGATACAAGAACTACTGATTTCAACTTCTTTGG
TGGACCTGATCAACTTCATCCAGGAGCTACTGACTTCAACTACTTCGG
Figure 1. Conservation of CpGs in the IFN-y proximal promoter and contiguous
transcribed region.
(A) Schematic showing the CpG locations in the IFN-y proximal promoter and
contiguous transcribed region. (B) A sequen ce alignment of t he IFN-y proximal
promoter and the contiguous transcribed region from mouse, rat, human, chimp and dog.
The CpGs located at the -190 and -53 positions relative to the transcription start site are
conserved between all shown species. The position of the proximal CREB/ATF and AP1
transcription factor binding site is also indicated.
66
Interestingly, one of the two CpGs that are conserved in the IFN-y promoter among
mouse, rat, dog, chimpanzee, and human resides in the proximal AP1 binding site (the
-53 CpG) (Figure 1), and methylation of this CpG has been shown to change factor
binding to this site, though it is unclear exactly what factors are affected by methylation
(Young et al., 1994). Considering that CpG methylation can inhibit the binding of some
AP1 family members to the promoter sequences of TrkA and neurotensin/neuromedin N
gene (Dong et al., 2000; Fujimoto et al., 2005), it is possible that CpG methylation plays
a direct role in regulating IFN-y transcription. However, the kinetics and the exact nature
and consequences of CpG methylation in the promoter on IFN-y transcription during TH2
polarization have yet to be elucidated.
We have here examined, in detail, the methylation changes that occur in the IFN-y
promoter during helper T cell polarization and how these changes affect transcription
factor binding and promoter function. We have shown that all CpGs within the IFN-y
proximal promoter are hypomethylated in naive CD4+ T cells and undergo progressive de
novo methylation during T. 2 polarization, with the CpG at the conserved -53 position
becoming methylated much more rapidly than the other CpGs. Methylation of the -53
CpG inhibits binding to the proximal AP1 CREB/ATF site by CREB and ATF2/c-Jun in
vitro. Correspondingly, the same factors bind to the hypomethylated IFN-y promoter in
TH1 cells but not the hypermethylated IFN-y promoter in TH2 cells.
Furthermore,
methylation of the -53 CpG alone is sufficient to inhibit the IFN-y promoter-driven
reporter gene expression in a TH1 cell line. Collectively, these findings suggest that the
rapid and site-specific methylation of the IFN-y promoter can immediately suppress
IFN-y transcription during TH2cell polarization by directly inhibiting transcription factor
binding.
67
Results
The IFN-y promoter is hypermethylated
in cells that do not express IFN-y
In mice, the IFN-y proximal promoter contains six CpG dinucleotides (Figure
1A), two of which, one at -53 position and the other at -190 position, are also conserved
in rat, dog, chimpanzee, and human (Figure 1B). Studies have shown that the IFN-y
proximal promoter is hypomethylated in IFN-y expressing TH1 cells and hypermethylated
in TH2cells, which do not express IFN-y (Young et al., 1994). This correlates with the
presumption that the promoters of silent genes are hypermethylated, while the promoters
of transcribed genes are hypomethylated (Jeltsch, 2002; Richards and Elgin, 2002). In
order to confirm that this correlation holds true for the IFN-y locus, we quantitatively
determined the methylation state of the IFN-y promoter in cells can do not express IFN-y.
To make a quantitative measurement of CpG methylation, we used sodium
bisulfite conversion of genomic DNA, followed by PCR amplification, cloning and
sequencing. Sodium bisulfite treatment, when followed by a deamination step, converts
unmethylatied cytosine to uracil. However, methylated cytosine residues are resistant to
bisulfite conversion.
During a subsequent PCR amplification, the uracil becomes
converted to thymadine. Therefore, by cloning and sequencing the PCR product we could
quantitatively determine the methylation percentage at a given CpG dinucleotide by
dividing the number of cytosine residues at a particular position by the total number of
sequenced clones.
In order to test the efficiency of conversion, we calculated the
frequency of conversion of C to T of cytosines that were not part of a CpG dinucleotide.
As non-CpG cytosines are almost never methylated in mammalian cells, we expected a
near 100% conversion of these bases. The conversion percentage we obtained was
greater than 99%, indicating the adequacy of this approach.
In order to determine the methylation state of the IFN-y promoter in cells that can
not transcribe IFN-y, genomic DNA was isolated from sorted B220+ B cells, kidney cells,
and heart cells, and was then subject to bisulfite conversion, PCR, cloning and
sequencing. As summarized in the schematic representation of representative sequences
and the total methylation percentages displayed in Figure 2, the kidney cells and heart
cells were almost completely methylated at both the IFN-y promoter and the proximal
68
Representative IFN-y promoter Sequences
A.
Kidney
Heart
AAA
VVW
la, Ift
AAA
WWW
AAA
A Ah Ah
AAA
AA
AAA
AAA
AAA
AhA
---
WW
AAA
- - A
- A
UVW
AA&
AAA
so$
lowW W
WWW
,a
II/WWW
AAA
###--WWW
AA&
WWW
WWW
V
A
WWW
Ah
WWW
WWW
AAA
W-4
VIVMr--VVV
AAA
AAA
B-cells
WV
A
W
---
V
AAAA
WaW
fvw
AA
AAA
-
-205 -190 -170
dA
A&
WmFvr--
VVW
VV
OUnmethylated CpG
B.
WWVW
W-~0M
W
s~~A'
WaN
A
A A
AAA
A
a
WWW
AAA A-Am
A
W W
AA
W
lAAA
IWWVWW
*Methylated
CpG
-53 -45 -34
16
96
Percentages of CpG methylation in cells that do not express IFN-y
Cell Type/Tissue
Kidney
Heart
B Cells
N
50
18
15
-205
96
100
80
-190
96
94
87
-170
100
100
73
-53
98
94
93
-45
96
94
87
-34
98
100
100
16
98
100
93
96
98
100
93
Figure 2. The IFN-y proximal promoter and contiguous transcribed region are
hypermethylated in cells that will never express IFN-y.
(A) Schematic representation of the methylation status of the CpGs within the IFN-y
proximal promoter and contiguous transcribed region of kidney, heart, and B cells. Data
shown is from 10 randomly selected clones, with the filled ovals representing methylated
CpGs and the empty ovals representing unmethylated CpGs. CpG methylation state was
determined by bisulfite conversion of genomic DNA, followed by PCR amplification,
cloning and sequencing. Heart and kidney cells were taken from Balb/c mice, while B
cells were isolated by FACS from the spleens of C57B16 mice. (B) The percentages of
CpG methylation at the IFN-y locus in heart, kidney and B cells. The number in the N
column represents the number of clones sequenced. Sequenced clones are from at least
two different PCR amplifications with two independently isolated DNA samples. The
other columns represent the percents of CpG methylation at the specific position at the
IFN-y locus in the indicated cell types or tissues as determined by bisulfite sequencing.
69
coding region (> 94%). While the IFN-y promoter and coding region was also highly
methylated in B cells, there was a slight, but significant (p < 0.001) reduction in the
methylation percentage in the promoter when compared to the kidney or heart cells.
While the physiological significance of this difference, if any, is unclear, IFN-y
expression by some human B cell subsets (Li et al., 1996) and human B cell lines (Pang
et al., 1992) has been demonstrated under certain conditions. It is therefore possible that
B cells may be slightly more predisposed towards IFN-y expression than heart or kidney
cells.
The IFN-y promoter is already hypomethylated in precursor T cells
It has been demonstrated that the IFN-y promoter is hypomethylated at the CpG
located at the -53 position relative to the transcriptional start site in IFN-y expressing THI
cell lines, and is hypermethylated at this position in TH2 cell lines that do not express
IFN-y (Young et al., 1994). In order to expand upon this observation, we determined the
methylation percentage for the six CpGs of the IFN-y proximal promoter and the two
most 5' CpGs of the transcribed region for the THI cell line AE7 and the TH2 cell line
D10 (Figure 3). The results we obtained confirmed the assumption that, in these cells,
the -53 CpG was representative of the other CpGs within proximal promoter. In the THI
cell line, each CpG within the IFN-y proximal promoter was completely unmethylated,
while the CpGs within the proximal transcribed region had only low levels of
methylation (
1%). In the TH2cell line, the CpGs within both the proximal promoter
and transcribed region were almost completely methylated ( 94%).
As has been
previously shown, the AE7 cells, but not the D10 cells, were able to rapidly express IFNy upon PMA and ionomycin stimulation (Figure 4). Thus, the CpG methylation patterns
within the TH1 clone AE7 and the TH2 clone DO10correlate with the IFN-y expression of
these cells.
While the fully differentiated THI and TH2 cell lines have IFN-y promoter
methylation patterns that correlate with gene expression, in naive CD4' T cells, from
which the THI and TH2 populations arise, the methylation status of the IFN-y promoter
has been reported as both hypermethylated and hypomethylated (Katamura et al., 1998;
Melvin et al., 1995; White et al., 2002; Winders et al., 2004).
To resolve this
70
Representative IFN-y promoter Sequences
A.
AE7 Cells
D10 Cells
(TH1)
(TH 2 )
we~~~~~~~.W __
~~~~~~~~AAnew
CellType
oo00
0N -205
0ooE4AAA
Ae -341
-190 -170AA
-53 VAA
-45
CpG
18
0
-205 -1 90-170
B.
0
-53
-45
18
94
100
WWW
WWW
W
WW 16
W
WW
AAAh
AAhhA[
96
A
Methylated
AUnmethylated
CpG
0
WAE7(TH1)
0
W0 0
-34
Percentagesof CpG methylationin THi
D 1 0(TH2)
''-WWW
100
16
and
1 00
6
11
100
100
96
TH
2
cell clones
1 00
1 00
Figure 3. The IFNy proximal promoter and contiguous
transcribed
AA
Ah
'A region are
-205IFN-y
-190 -170
-53and
-45 Din
-34 cells. regosent
16 number
96
diffeigurent
PCR
amplifications.
promoter
ncolumnsre
CpG
B.hypomethylated
Percentagmethylation
at the
locusand
inThe
A
E7
The
in the N of
columnes
hypomethylated
hypermethylated
D1O
cells.
in AE7
AE7 cells
cells
and
hypermethylated
in D10
cells.the percents
represents
thype
number
of clones
sequenced.
are 0from
at leastIFN-y
(A)
Schematic
representation
methylation
status
(A)
Schematic
representation
of the
tWW
he0CpGs
wWithin
AE7
(TH1)
18
0 methylation
0 Sequenced
0status
1 1two96
OW0ofclones
W6 the IFN-y
D 10 (TH2)
18 94
100 100 100 100 100 100 100
the TH
THi cell line AE7 and the
proximal promoter and contiguous transcribed region of the
TH2
Data
randomly
clones,
with the filledCpGs.
ovals
TH2cell line D1O.
D0.
shown
is
from
10
selected
methylatedOUrepresenting
CpGs and the empty ovals representing unmethylated
CpG m ethylation state was determined by bisulfite conversion of genomic DNA,
followed by PCR amplification, cloning and sequencing. (B) The percentages of CpG
representing
methylationed
CpGs a
the emptIFN-y
locuvalsrepresenting unmethyldcated
cell types or tissues
pas
determined
by bisulfitate
was determined
by sequencing.
bg.sulfite conversion
of genomic ofDNAg.
followed
by PCR
amplification,
cloning and
(B) The percentages
CpG
methylation
at number
the IFN-y
A E7 and D0
cells. The
number
in theatNleast
column
represents
the
of locus
clonesinThe
sequenced.
Sequenced
clones
are from
two
different PCR
amplifications.
other columns
represent
the
percents
of CpG
methylation at the specific position at the IFN-y locus in the indicated cell types or tissues
as determined by bsulfite sequencing.
71
IFN-y
D 1 hour of PMA and
lonomycin Stimulation
[~J No Stimulation.
IIlsotype
Control
Figure 4. The T HI cell line AE7 expresses IFN-y when stimulated, but the T H2 cell
line D10 does not.
FACS histograms showing the IFN-y expression of AE7 and D 10 cells. AE7 and D 10
cells were either stimulated with PMA and i onomycin for 2 hours, or left without
stimulation, and IFN-y expression was measured by cytokine recapture FACS.
Stimulated cells were also stained with an isotype control antibody.
The sh aded
histogram represents the isotype control, the dashed histogram represents cells without
stimulation and the solid line represents cells with stimulation.
72
inconsistency, we measured the extent of methylation of all six CpGs in the proximal
promoter, as well as the two CpGs in the transcribed region immediately following the
promoter, in purified naive CD4+ T cells (Figure 5). We found that in DNA isolated from
freshly purified naive CD4+ T cells, which do not transcribe IFN-y, the IFN-y promoter
was hypomethylated ( 9% methylation at any CpG), while the transcribed region of the
gene was hypermethylated (
85% methylation at both CpGs).
These results are
consistent with and further extend those reported by Winders et al.
Like naive CD4* T cells, naive CD8+ T cells do not express IFN-y, but have the
capacity to express IFN-y following activation (Grayson et al., 2001). To determine
whether these cells have the same methylation pattern at the IFN-y promoter as the naive
CD4+ T cells, we purified fresh naive CD8+ T cells and analyzed their methylation state
(Figure 5). Like naive CD4+ T cells, in this cell population the IFN-y promoter was
hypomethylated ( 4% methylation at any CpG), while the transcribed region of the gene
was hypermethylated ( 92% methylation at both CpGs).
CD4+ T cells and CD8+ T cells are derived from a common progenitor within the
thymus, while B cells and T cells are derived from an earlier progenitor within the bone
marrow.
As the IFN-y proximal promoter is hypermethylated
in B cells, but
hypomethylated in naive CD4+ and CD8+ T cells, it is possible that demethylation of the
IFN-y promoter takes place during thymic development.
We investigated the stages
during T cell development in which the IFN-y promoter first becomes hypomethylated.
To accomplish this, CD4-CD8- (double negative or DN), CD4+CD8+ (double positive or
D)P), CD4 + (single positive or SP), and CD8 + SP thymocytes were purified by FACS and
assayed for methylation at the IFN-y locus. In all four thymocyte populations, the IFN-y
promoter was hypomethylated (< 5% methylation at any CpG) while the transcribed
region was hypermethylated ( 96% methylation at both CpGs) (Figure 6), showing the
same IFN-y promoter methylation phenotype as naive CD4+ and CD8+ T cells that were
isolated from spleens and lymph nodes. To further determine the developmental stages
when IFN-y promoter becomes hypomethylated, and to further correlate IFN-y expression
with IFN-y locus methylation, we assayed the methylation status of freshly purified
natural killer (NK) cells. Like T cells, NK cells are differentiated from DN thymocytes,
73
Representative IFN-y promoter Sequences
A.
Nai'Ve CD4+ T cells
NaVe CD8+ T cells
0
0~~ooE
QOO-G&--(
ooo-oo^
ooo=oo0L
*oo-ooQ-z
oo00ooL
.oo
QO-OO7
ooo-.o-'
C.oE
ooo~rooo.5L
oo00_00~E
oooo~~E
OOO-OOOUnmethylated
-205 -190 -170
B.
CpG
OMethylated
-53 -45 -34
CpG
16
96
Percentages of CpG methylation in naive T cells
Cell Type
Naive CD4 + T ccells
Na've CD8 + T ccells
N
34
26
-205
3
4
-190
9
0
-170
3
4
-
-53
3
4
-45
0
4
-34
0
0
16
85
92
96
97
100
-
Figure 5. The IFN-y proximal promoter is hypomethylated and contiguous
transcribed region is hypermethylated in naive T cells.
(A) Schematic representation of the methylation status of t he CpGs w ithin the IFN-y
proximal promoter and contiguous transcribed region of naive CD4+ and CD8+ T cells.
Data shown i s from 10 randomly selected clones, with the filled ovals representing
methylated CpGs and the empty ovals representing unmethylated CpGs. CpG
methylation state was determined by bisulfite conversion of genomic DNA, followed by
PCR amplification, cloning and sequencing. Naive CD4+ and Na've CD8+ were isolated
from the spleens and lymph nodes of Balb/c mice by FACS. (B) The percentages of CpG
methylation at the IFN-y locus in different thymocyte populations.
The number in the N
column represents the number of clones sequenced. Sequenced clones are from at least
three different PCR amplifications with at least two independently isolated DNA
samples. The other columns represent the percents of CpG methylation at the specific
position at the IFN-y locus in the indicated cell types or tissues as determined by bisulfite
sequencing.
74
Figure 6
Representative IFN-y promoter Sequences
A.
Double Positive
Thymocytes
Double Negative
Thymocytes
ooo-oo4
4
ooo-Go-
oo.-ooo- 4
CD8 Thymocytes
4
ooo-ooo
Qoo0QCF
+
ooo-o~o-'4
0 0~~oo1 ooo-ooo4
ooo-ooo
14
E4
ooo-ooo
,o~oo~E
Qo0ooCL
ooo-oo'4
ooo-&oo-i~
0o.L
oo0
QO00
OUG0~
E
Qoo~o-oooE4
(X
QOO-QO(
oo-o0-'
ooo-ooo-'4
QQCE
ooo0oQ>E
Qo-Oo(
oo-ooC
oo00
ooE
oo00
o&E
oo0
ooo-~ooo E4
CpG f Methylated CpG
(nUnmethylated
-205 -190 -170
B.
0o&E6
-53
-45
16
-34
96
Percentages of CpG methylation in Thymocyte Populations
Cell Type/Tissue
N -205 -190 -170 -53
-45
-34
16
96
T
DN Thymocytes
DP Thymocytes
CD4 Thymocytes
CD8 Thymocytes
=
20
29
25
23
0
0
0
0
0
0
4
0
5
0
4
0
0
0
0
4
0
3
0
0
5
3
4
4
100
97
96
100
100
100
100
100
75
Figure 6. The IFN-y proximal promoter is hypomethylated and contiguous
transcribed region is hypermethylated in thymocytes.
(A) Schematic representation of the methylation status of t he CpGs w ithin the IFN-y
proximal promoter and contiguous transcribed region of different thymocyte populations.
Data shown i s from 10 randomly selected clones, with the filled ovals representing
methylated CpGs and the empty ovals representing unmethylated CpGs. CpG
methylation state was determined by bisulfite conversion of genomic DNA, followed by
PCR amplification, cloning and sequencing. The double positive and single positive
thymocytes were isolated from the thymuses of C57BL6 mice by FACS, while the double
negative thymocytes were isolated from the thymuses of RAG 1 deficient C57BL6 mice
by FACS. (B) The percentages of CpG methylation at the IFN-y locus in different
thymocyte populations. The number in the N column represents the number of clones
sequenced. Sequenced clones are from at least two different PCR amplifications with
two independently isolated DNA samples. The other columns represent the percents of
CpG methylation at the specific position at the IFN-y locus in the indicated cell types or
tissues as determined by bisulfite sequencing.
76
but unlike naive T cells, NK cells constitutively transcribe IFN-y (Stetson et al., 2003).
Within this population of cells, all CpGs in both the promoter and the transcribed region
were completely demethylated (Figure 7). These results suggest that the IFN-y promoter
becomes hypomethylated after lymphoid lineage commitment but before bifurcation
between T cell and NK cell development in the thymus.
Hypomethylation of the IFN-y promoter is correlated with the cells' potential to
express IFN-y
The above results indicate that hypomethylation of the IFN-y promoter is
correlated with the ability of a cell to differentiate into an IFN-y expressing cell, rather
than with the cell's actual expression of the gene. To further investigate this possibility,
we assayed the methylation status of the IFN-y locus in CD4 + T cells after THI1 and TH2
polarization. Naive CD4+ T cells were purified from Balb/c mice and activated by
exposure to plate-bound anti-CD3 and anti-CD28, and recombinant IL-2.
TO
polarization was induced by addition of IL-12 and anti-IL-4 antibody to the culture, while
TH2 polarization was induced by addition of IL-4 and anti-IFN-y antibody to the culture.
Six days later, methylation was assayed in polarized cells. This in vitro polarization
results in the expected difference cytokine expression between the cell populations, as
stimulation rapidly induces IFN-y but not IL-4 expression in the THl polarized cells, and
IL-4 but not IFN-y expression in the TH2 polarized cells (Figure 8).
Following culture under T
hypomethylated
conditions,
the IFN-y promoter remained
(< 11% methylation at each CpG), while the transcribed region
underwent a level of demethylation (51% and 76% methylation at the +16 and +96
position CpGs respectively) as compared to naive CD4+ T cells (p < 0.001) (Figure 9). It
is notable that, while the CpG located at the +16 position has undergone a greater degree
of demethylation than the CpG located at the +96 position, there are clones in which the
+16 position CpG is methylated, but the +96 position CpG is demethylated.
This
suggests that the +16 position CpG is more prone to demethylation than the +96 position
CpG, but that the demethylation is not progressive from a 5' to a 3' direction.
When cells were cultured under TH2polarizing conditions, during which IL-4 but
not IFN-y was expressed (Figure 8), the CpGs within the IFN-y promoter became
77
Representative IFN-y promoter Sequences
A.
Natural Killer Cells
Day 6 THO Cells
QOOOGCE
QOOOQ-(
ooooo-(
QOO-G#O-(
00-GO-0
oo00o( EI
QOO(GO00E
00_0
00~OOE
QOOOO-(
GOO-QOZ
QO00QOE
QO00OOE
QOOCOGe
QOOOQ-(
QOO0OO(50
QO-G
_~
QOO~OQE
OUnmethylated CpG *Methylated
-205 -190 -170
B.
-53 -45 -34
CpG
16
96
Perccentages of CpG methylation in natural killer cells
Cell Type
Natural Killer Cells
14
Day 6 THO Cells
31
N
-205 -190
0
0
0
10
-170
0
-53
-45
-34
0
0
0
10
13
0
7
16
0
87
96
0
87
Figure 7. The IFN-y proximal promoter is hypomethylated in natural killer (NK)
cells and unpolarized, activated helper T cells (THO cells), and contiguous
transcribed region is hypomethylated in natural killer cells, but hypermethylated
THO cells.
(A) Schematic representation of the methylation status of t he CpGs w ithin the IFN-y
proximal promoter and contiguous transcribed region of natural killer cells and THOcells.
Data shown i s from 10 randomly selected clones, with the filled ovals representing
methylated CpGs and the empty ovals representing unmethylated CpGs. CpG
methylation state was determined by bisulfite conversion of genomic DNA, followed by
PCR amplification, cloning and sequencing. The NK cells were isolated from the spleens
of RAGI deficient C57BL6 mice by FACS, while THOcells were generated by activating
FACS purified naive Balb/c CD4+ T cells with plate-bound anti-CD3 and anti-CD28
antibodies for 6 days in the presence of neutralizing anti-IFN-y and anti-IL-4 antibodies.
(B) The percentages of CpG methylation at the IFN-y locus in NK and THOcells. The
number in the N column represents the number of clones sequenced. Sequenced clones
are from at least two different PCR amplifications with two independently isolated DNA
samples. The other columns represent the percents of CpG methylation at the specific
position at the IFN-y locus in the indicated cell types or tissues as determined by bisulfite
sequencing.
78
IFN-y
IL-4
D 2 hour of PMA and lonomycin
IIlsotype
stimulation
Control.
Figure 8. In vitro polarized T HI cells express IFN-y, but not IL-4, while in vitro
polarized T H2 cells express IL-4, but not IFN-y.
IFN-y and IL-4 expression by polarized T H I and TII2 cells. Day 6 polarized Till and TII2
cells were stimulated with PMA and ionomycin for 3 hours and cytokine expression was
measured by cytokine recapture FACS. The open histogram represents anti-IFN-y or
anti-IL-4 staining, while the shaded histogram represents an isotype control.
79
Figure 9
Day 6 TH1 Cell IFN-y promoter Sequences
A.
*o Ore E4
QOOOQCF
www
oo00CoE
QO-0O'
QO-GO0
00(\.
000-0
ooo0(oo
ooo0oCFE
QO00QOE
QOO00~E
E
oo00ooE
oooCooE
QO-OO-(
000~OOL
0
Unmethylated CpG 0
Methylated CpG
Q0O(XGGQ~L~
_QOQOOEA
00 -GOQO-GO-0-
UU-y
-205 -190 -170
-53
-45
16
-34
96
B.
Percentages of CpG methylation in Day 6 TH1 Cells
Cell Type
N
Day 6 TH1 Cells
45
-205
7
-190
4
-170
7
-53
11
-45
4
-34
9
16
51
96
76
80
Figure 9. The IFN-y proximal promoter is hypomethylated and the contiguous
transcribed region is partially demethylated
in Day 6 TH1 cells.
(A) Schematic representation of the methylation status of the CpGs within the IFN-y
proximal promoter and contiguous transcribed region of day 6 TH cells. Data shown is
from all sequenced clones, with the filled ovals representing methylated CpGs and the
empty ovals representing unmethylated CpGs. CpG methylation state was determined by
bisulfite conversion of genomic DNA, followed by PCR amplification, cloning and
sequencing. TH1 cells were generated by activating FACS purified naive Balb/c CD4+ T
cells with plate-bound anti-CD3 and anti-CD28 antibodies for 6 days in the presence of
recombinant IL- 12 protein and neutralizing anti-IL-4 antibodies. (B) The percentages of
CpG methylation at the IFN-y locus day 6 TH cells. The number in the N column
represents the number of clones sequenced. Sequenced clones are from four different
PCR amplifications with three independently isolated DNA samples. The other columns
represent the percents of CpG methylation at the specific position at the IFN-y locus in
the indicated cell types or tissues as determined by bisulfite sequencing.
81
hypermethylated as compared to that in naive CD4+ T cells (p < 0.001), while the
transcribed region remained hypermethylated (Figure 10). In particular, the -53 CpG
became significantly more methylated than the other CpGs within the promoter (70%
versus 13-32%, p
0.001).
There were specific clones in which there was not
methylation of the -53 position CpG, but that there was methylation at other CpGs. This
indicates that, while the -53 CpG is more prone to becoming methylated than the other
proximal promoter CpGs during TH2 polarization, de novo methylation at the -53
position is not required for methylation of the other CpGs. When the cells were activated
in the presence of IFN-y and IL-4 neutralizing anti-bodies, but in the absence of
polarizing cytokines (THO),there was no detectable change in methylation of the IFN-y
promoter as compared to naive CD4+ T cells (Figure 7).
We also assayed the methylation status of the IFN-y locus in effector and memory
CD8+T cells, both of which transcribe IFN-y (Grayson et al., 2001). The effector CD8+T
cells were generated in vitro by activating purified naive CD8+ T cells with anti-CD3 and
anti-CD28 for three days. To generate memory CD8+T cells, sorted naive CD8' T cells
from a mouse expressing a transgenic 2C T cell receptor were injected into RAG
deficient mice. The mice were immunized with the 2C cognate peptide SIYRYYGL and
complete Freund's adjuvant (CFA), and, after 6 months, CD44 high CD8+ T cells were
sorted from lymph nodes and spleens.
As shown in Figure 11, the IFN-y promoter
remained hypomethylated in these CD8+T cells. Notably, the CpGs at the +16 and +96
positions were more hypomethylated in memory CD8+T cells (31% and 69% methylation
respectively)
than in naive CD8+ T cells (92% and 100% methylation
respectively)
(p <
0.001). These results further support the notion that hypomethylation of the IFN-y
promoter is correlated with the cells' potential to express IFN-y. In addition,
hypomethylation of the transcribed region in NK cells, THI cells, and memory CD8+ T
cells and hypermethylation of the promoter in TH2 cells suggest dynamic changes in the
methylation status of the IFN-y locus in association with the cell's history of IFN-y
transcription.
82
Figure 10
Day 6 TH 2 Cell IFN-y promoter Sequences
A.
QO-O-^
oo.rmoyE
GOwOOeE
Q00QE
0
0~~&OE
oo00
1
000~n~E
000-
B
Qo00c(L~
ew
oo00
ooo ooE
0
Unmethylated CpG
*
o.0E
Methylated CpG
oo00ooE
oo00ooE
QO0
0Q~E
oo00
CoE
oo000o~E
oo&oo(E
000
00~4
000_400s1
oo
OO00G-~
&o4
oo00
o.
vvv
-205 -190 -170
-53
-45
-34
16
96
B.
Percentages of CpG methylation in Day 6 TH1 Cells
CellType
Day 6 TH2 Cells
N
47
-205 -190 -170
26
28
32
-53
-45
-34
16
96
70
13
21
96
98
-
83
Figure 10. IFN-y proximal promoter has undergone de novo methylation, and the
contiguous transcribed region is hypermethylated
in Day 6 TH2 cells.
(A) Schematic representation of the methylation status of the CpGs within the IFN-y
proximal promoter and contiguous transcribed region of day 6 TH2 cells. Data shown is
from all sequenced clones, with the filled ovals representing methylated CpGs and the
empty ovals representing unmethylated CpGs. CpG methylation state was determined by
bisulfite conversion of genomic DNA, followed by PCR amplification, cloning and
sequencing. TH2cells were generated by activating FACS purified naive Balb/c CD4 + T
cells with plate-bound anti-CD3 and anti-CD28 antibodies for 6 days in the presence of
recombinant IL-4 protein and neutralizing anti-IL-12 antibodies. (B) The percentages of
CpG methylation at the IFN-y locus day 6 TH2cells. The number in the N column
represents the number of clones sequenced. Sequenced clones are from four different
PCR amplifications with three independently isolated DNA samples. The other columns
represent the percents of CpG methylation at the specific position at the FN-y locus in
the indicated cell types or tissues as determined by bisulfite sequencing.
84
Representative IFN-y promoter Sequences
00(~7-
Effector CD8 + T cells
A.
Memory CD8+ T cells
QOO-004
4
oo-0o'
oo-ooocy
000
A&&
o0
&
OOO-GGOd'p
000-00
Q0C8OOnmyl4
oo00ooCL
oooooJ.
Qo0 0~o1
OUnmethylated
CpG
OMethylated
CpG
Percentages of CpG methylation in Effector and Memory CD8+ T cells
B.
Cell Type
Effector CD8+ T cells
Memory CD8+ T cells
N
-205 -190 -170
0
0
0
-
23
16
6
6
6
-53
0
0
-45
-34
16
96
0
0
0
0
65
31
91
69
-
Figure 11. The IFN-y proximal promoter is hypomethylated and the contiguous
transcribed region has undergone demethylation in effector and memory CD8+ T
cells.
(A) Schematic representation of the methylation status of t he CpGs w ithin the IFN-y
proximal promoter and contiguous transcribed region of effector and memory CD8+ T
cells. Data shown is from 10 randomly selected clones, with the filled ovals representing
methylated CpGs and the empty ovals representing unmethylated CpGs. CpG
methylation state was determined by bisulfite conversion of genomic DNA, followed by
PCR amplification, cloning and sequencing. Effector CD8+ T cells were generated by
stimulating FACS purified naive C57B16 CD8+ T cells with anti-CD3 and anti-CD28
antibodies for 3 days. Memory CD8+ T cells were generated by intravenously adoptively
transferring naive CD8+ T cells from RAG-1 deficient, transgenic 2C TCR expressing
C57B16 mice into RAG-1 deficient C57B16 mice. These mice were immunized with
SIYRYYGL peptide in CFA. Approximately 6 months later, memory CD8 T cells were
purified from these mice by FACS. (B) The percentages of CpG methylation at the IFNy locus in different thymocyte populations. The number in the N column represents the
number of clones sequenced. Effector CD8+ T ce 11sequenced clones are from three
different PCR amplifications with two independently isolated DNA, while the memory
CD8 T cell clones are from two different PCR amplifications, but a s ingle DNA
isolation. The other columns represent the percents of CpG methylation at the specific
position at the IFN-y locus in the indicated cell types or tissues as determined by bisulfite
sequencing.
85
The -53 CpG in the IFN-y promoter is rapidly methylated during TH2 cell
polarization
Increased level of methylation of the IFN-y promoter in day 6-polarized TH2 cells
suggest that the IFN-y promoter can become de novo methylated during TH2 cell
polarization. To unequivocally demonstrate this event, we examined the methylation
changes at the IFN-y locus during TH1 and TH2 polarization over 19 and 25 days
respectively. As before, naive CD4+ T cells were purified and activated under either TH1
or TH2 polarizing conditions.
Polarizing conditions were maintained, and cells were
restimulated with anti-CD3 and anti-CD28 every 6 to 7 days. DNA was isolated from the
T cells for methylation analysis on day 3 and prior to each restimulation.
T cells cultured under both TH1 and TH2 polarizing conditions underwent
progressive methylation changes at the IFN-y locus.
In T cells cultured under TH1
conditions, the promoter remained hypomethylated and the transcribed region became
increasingly demethylated from day 3 to day 19 (Figures 12 and 13). At all time points,
the CpG at the +16 position was less methylated than the CpG at the +96 position,
suggesting that the +16 position CpG is more prone to demethylation than the +96
position CpG. There were again instances in which the +16 position CpG remained
methylated but the +96 position CpG was demethylated, confirming that progressive 5' to
3' demethylation is unlikely. In T cells polarized under TH2 conditions, in contrast, the
promoter became increasingly methylated from day 3 to day 25, while the transcribed
region remained hypermethylated (Figures 14 and 15). These patterns in methylation
change are consistent with the observation that the long-term TH1 polarized cell line AE7
is almost completely hypomethylated at the IFN-y locus, while the long-term TH2
polarized cell line D10 is almost completely hypermethylated at the IFN-y locus (Figure
3). It is notable that during the TH2 polarization the methylation rate varied from CpG to
CpG, with the CpG located at the -53 position consistently becoming methylated earlier
than other CpGs in the promoter. As the -53 CpG is evolutionarily conserved (Figure 1),
and the -53 CpG resides in the proximal AP1 binding site of the IFN-y promoter, it is
possible that the rapid de novo methylation of the -53 CpG may help to promote TH2 cell
polarization by inhibiting IFN-y transcription.
86
TH1 Polarization IFN-y promoter Representative Sequences
Figure 12
A.
Day3
0
Day6
Day 13
Qoo-0oZ
0-OQ~
Qdd SPAoQE
oo-oow
oooIooKE
oooooo}0i
QO-GO(
QOO_
ooo0G(0E
QO-OO-0-
o-C
000o
oo00
0
ooco&E
000G&-
4
_O_ _
0 0~~G&L
QOOGG-0
QQE
OMethylated CpG
Unmethylated CpG
Day 19
ooo-~ooo-E4
oo-ooo-Io
ooo-ooo-E4
QO00GGE6
QO0 GGC1
-53 -45 -34
-205-190-170
16
96
B.
Percentages of CpG methylation in TH1 Cells
Cell Type
Day 3 TH1 Cells
Day 6 TH1 Cells
Day 13 TH1 Cells
Day 19TH1 Cells
N
w
-
-205
-190
-170
-53
-45
-34
16
96
- -
32
3
9
3
3
0
0
85
97
45
19
21
7
0
0
4
5
5
7
0
5
11
11
0
4
5
0
9
0
0
51
37
24
76
53
33
87
Figure 12. The IFN-y proximal promoter remains hypomethylated and contiguous
transcribed region becomes hypomethylated during TH polarization.
(A) Schematic representation of the methylation status of t he CpGs w ithin the IFN-y
proximal promoter and contiguous transcribed region of TH cells isolated at day 6, 13
and 19. Data shown is from 10 randomly selected clones each, with the filled ovals
representing methylated CpGs and the empty ovals representing unmethylated CpGs.
CpG m ethylation state was determined by bisulfite conversion of genomic DNA,
followed by PCR amplification, cloning and sequencing. Polarized TH cells were
stimulated every 6-7 days and w ere harvested for methylation analysis just prior to
restimulation. (B) The percentages of CpG methylation at the IFN-y locus on different
days during TH polarization. The number in the N column represents the number of
clones sequenced. Sequenced clones are from at least two different PCR amplifications
with two independently isolated DNA samples. The other columns represent the percents
of CpG methylation at the specific position at the IFN-y locus in the indicated cell types
or tissues as determined by bisulfite sequencing. These data are also shown in Figure 13.
88
IFN-y Methylation Changes
During TH1 Polarization
100
-
* -205 CPG
80
*
c
o
--- 170 CpG
60
ra)
X -53 CpG
N- -47 CpG
4-_
40
I * -34 CPG
0
I÷ +16 CpG
-+96 CpG
20
0
-190 CpG
I
-
- -
M
0
MfttI
!
_
10
5
I
15
20
Time (Days)
Figure 13. The IFN-y proximal promoter remains hypomethylated and contiguous
transcribed region becomes hypomethylated during TH polarization.
A graph showing the changes in the levels of methylation at different CpGs in the IFN-y
locus during the course of TH1 polarization. Polarized TH1 cells were stimulated every 67 days and were harvested for methylation analysis just prior to restimulation. These data
are also shown in Figure 12.
89
TH2 Polarization Representative IFN-y promoter Sequences
Figure 14
Dav 3
A.
Day 6
Day 13
_ww
Co-oQOA
-
A_-'
0 *Mehlae
Une0yatd0p
Day 2
0
QO&i4QiE
Day 19
AAA
CpG
oo00_00.
_ww
oo00_o~E
QOOPLEAOAs4
QO00
GOE
000_001.*Q *r
1^
B.
Percentages of CpG methylation in TH2 Cells
Cell Type
Day 3 TH2 Cells
Day 6 TH2 Cells
Day 13 TH2 Cells
Day 19 TH2 Cells
Day 25 TH2Cells
-
N
- -205
28
47
29
30
28
25
32
42
47
64
|
-190 -170
25
14
26
28
63
29
60
30
82
32
-53
43
70
92
97
100
-45
0
13
21
23
29
-34
17
21
4
17
25
16
86
96
100
97
100
96
96
98
100
100
96
90
Figure 14. The IFN-y proximal promoter becomes hypermethylated and contiguous
transcribed region remains hypermethylated during TH2 polarization.
(A) Schematic representation of the methylation status of the CpGs within the IFN-y
proximal promoter and contiguous transcribed region of TH2 cells isolated at day 6, 13,
19 and 25. Data shown is from 10 randomly selected clones each, with the filled ovals
representing methylated CpGs and the empty ovals representing unmethylated CpGs.
CpG methylation state was determined by bisulfite conversion of genomic DNA,
followed by PCR amplification, cloning and sequencing. Polarized TH2 cells were
stimulated every 6-7 days and were harvested for methylation analysis just prior to
restimulation. (B) The percentages of CpG methylation at the IFN-y locus on different
days during TH2 polarization. The number in the N column represents the number of
clones sequenced. Sequenced clones are from at least two different PCR amplifications
with two independently isolated DNA samples. The other columns represent the percents
of CpG methylation at the specific position at the IFN-y locus in the indicated cell types
or tissues as determined by bisulfite sequencing. These data are also shown in Figure 15.
91
IFN-y Methylation Changes
During TH2 Polarization
100
C
60
* -205 CpG
---- 190 CpG
&-170 CpG
40
---- -53 CpG
X -47 CpG
80
O
o
ca)
-0
- *-34 CpG
A
1~~~~~~~~~~~~~~~~~M
20
..
0
0
5
10
15
Time (Days)
20
+16 CpG
+96 CpG
25
Figure 15. The IFNy proximal promoter becomes hypermethylated and contiguous
transcribed region remains hypermethylated during TH2 polarization.
A graph showing the changes in the levels of methylation at different CpGs in the IFN-y
locus during the course of TH2 polarization. Polarized TH2 cells were stimulated every 67 days and were harvested for methylation analysis just prior to restimulation. These data
are also shown in Figure 12.
92
Methylation of the -53 CpG impairs ATF2/c-Jun and CREB binding in vitro
We next investigated the effect of -53 CpG methylation on transcription factor
binding by electrophoretic mobility shift assays (EMSA). In these studies, we used
nuclear extracts from the mouse TH1 cell line AE7 and oligonucleotide probes
corresponding to the proximal AP1 site (-62 through -32) of the IFN-y promoter that
were either unmethylated, or methylated at the -53 CpG, the -47 CpG, the -34 CpG, or
methylated at all three CpGs simultaneously.
When an EMSA was done with the
unmethylated oligonucleotide probe, there were two major shifted bands (Figure 16),
indicating factor binding. Methylation at either the -47 CpG or the -34 CpG had no
effect on the formation of these bands. In contrast, the same sequence methylated at only
the -53 position or at all three positions abolished the formation of the EMSA complexes.
To confirm the observed results, we carried out competition assays in which
increasing amounts of unlabeled competitor oligonucleotides were added to a fixed
amount of labeled, unmethylated probe during the incubation step of EMSA assays. As
expected, formation of the two complexes was effectively inhibited by an excess of the
unlabeled, unmethylated probe (Figure 17). Oligonucleotides that were methylated at
either the -45 or -34 CpG also inhibited the formation of the two complexes to a similar
degree. In contrast, oligonucleotides that were methylated at the -53 position, or at all
three CpGs, were much less effective, but still more effective than non-specific NFAT
oligonucleotide, in inhibiting the complex formation (Figure 17). Based on the relative
intensities of the shifted bands, -53 CpG methylation reduced factor binding between 10fold and 100-fold. These results show that methylation at the -47 or -34 CpGs does not
affect in vitro factor binding, while methylation at the -53 CpG significantly reduces in
vitro factor binding.
It has been shown that various members of the AP1 and CREB/ATF families of
transcription factors are able to bind the proximal AP1 site (Cippitelli et al., 1995; Penix
et al., 1996; Yano et al., 2003). We next wished to determine the identity of factors
present in the complexes we observed in the above EMSA assays. In order to accomplish
this, we next determined the composition of transcription factors in the two complexes by
inclusion of specific antibodies in the binding reaction of the EMSA assay. Antibodies
specific for FosB, c-Fos, and JunB did not significantly affect the two shifted bands
93
Figure 16. Inhibition of factor binding by methylation of the -53 position CpG.
An EMSA using probes corresponding to the proximal API CREB/ATF site of the IFN-y
promoter that has been methylated CpG -53, CpG -45, CpG -37, or at all three CpGs on
the same probe, and nuclear extracts from the TH 1 cell line AE7. Data shown are
representative of greater than three experiments.
94
Probe :Competitor Ratio
~..~ ~..~.~
~.
No
Compo
No Me
-53 Me
-45 Me
~
.~ .~ ~
~.~.~.~.
-34 Me
~
.~.~.~
~. ~.
Trip Me
~
NFAT
Competitor Methylation
Figure 17. Methylation of the -53 position CpG inhibits factor binding by greater
than 10 fold.
An EMSA in the presence of competitor oligonucleotides was performed using the AE7
nuclear extract and labeled, unmethylated probe corresponding to the proximal API
CREB/ATF site of the IFN-y promoter. The EMSA was carried out in the presence of
increasing amounts of unlabeled methylated or unmethylated or non-specific (NFAT)
competitor oligonucleotides. The first lane was from the incubation without competitor
oligonucleotide. Data shown are representative of greater than three experiments.
95
(Figure 18A). In contrast, antibodies specific for c-Jun supershifted the upper complex,
and those specific for ATF2 diminished its formation. Furthermore, antibodies specific
for CREB or both ATF1 and CREB supershifted the lower complex.
In order to test whether the components required for the formation of the two
major complexes were present in TH2 cells, as well as THI cells, we next repeated the
above supershift EMSA experiments using nuclear extracts from the mouse TH2 cell line
D10, but with the same probes and antibodies as before. When the TH2 nuclear extract
was used, the same two complexes were detected and the same effects were observed
with the various antibodies (Figure 18B). Thus, the upper complex contains ATF2/c-Jun
and the lower complex contains CREB (and possibly ATF1), and these factors are present
in the nucleus of both
THI1
and TH2 cells.
CREB, ATF2 and c-Jun are bound to the IFN-y promoter in THI but not TH2cells
To assess CREB and ATF2/c-Jun binding to the IFN-y promoter in vivo, we
performed chromatin immunoprecipitation (ChIP) assays using the THI cell line AE7 and
the TH2 cell line D10. As described above, the IFN-y promoter is nearly completely
unmethylated in AE7 cells and nearly completely methylated in D10 cells (Figure 3), and
therefore should give a clear indication as to any correlation between promoter
methylationand factor binding in vivo.
Cells were treated with formaldehyde to cross-
link proteins to DNA, and lysed. Sonication was used to fragment the DNA to sizes
between 200 bp and 500 bp. This was followed by immunoprecipitation with various
antibodies. The presence of IFN-y promoter sequence in the precipitates was then
determined by PCR.
As expected, PCR signal of the IFN-y promoter was detected in DNA isolated
from pre-precipitates, but not in precipitates where no antibody was added in both the
THI
and the
TH2
samples (Figure 19). In DNA from TH1 precipitates, the same size PCR
product was detected when antibodies specific for FosB, JunB, c-Jun, both ATF1 and
CREB, ATF2, and CREB alone were used, but not when anti-c-Fos was used (Figure 19),
indicating binding of FosB, JunB, c-Jun, ATF2 and CREB to the IFN-y promoter in these
cells. In DNA from TH2 precipitates, the PCR product was detected when antibodies
specific for FosB, c-Fos, JunB, and both ATF1 and CREB was used, but not when
96
Antibody
A.
~
"$)0
~o
~
~o
Q.
0'"
~
$>
~
~o
~
':J.::5
G
s
;S
G
~'
~~
" cr
~CV
"
~
~
~
w
u
....
W
~
~~
W1
AE7 (T H 1) Nuclear Extract
Antibody
B.
~
"$)0
~o
~
Q.~o
$>
~
~o
~
G
0'"
~
~
;SS
G
~'
"
~
~~
" cr
~CV
~
~
010 (TH2) Nuclear Extract
Figure 18. c-Jun, ATF2 and CREB from both T HI and T H2 cell lines are able to bind
to the IFN-y proximal API CREB/ATF binding site.
EMSA assays done using unmethylated probe corresponding to the proximal AP I
CREB/ ATF binding site of the IFN-y promoter and nuclear extracts from the T HI cell line
AE7 (A) or the TH2 cell line D 10(8). The assays were carried out in the presence of the
indicated specific antibodies. The EMSA with labeled, methylated API probe serve as
controls. The anti-ATFI antibody cross-reacts with CRE8. Data shown are
representative of greater than three experiments.
97
Antibody
Promoter
Intron 1
Promoter
Intron 1
Figure
19. Transcription
factor binding to the IFN-y promoter in T HI and T H2 cell
lines.
Chip assay of the indicated API and CREB/ATF
family members binding within the
IFN-y promoter in AE7 and D 10 cell lines. DNA and proteins were cross-linked
in AE7
or D I0 cells, followed by sonication and precipitation
with the indicated antibody.
The
cross-linking
was reversed,
and t he presence of I FN-y DNA in the precipitates
was
determined by PCR. PCR assays of IFN-y intron I were used to demonstrate
specificity
of the immunoprecipitation.
Representative
data from one
of two independent
experiments is shown.
98
antibodies specific for c-Jun, ATF2 and CREB alone were used (Figure 19), indicating
the binding of FosB, c-Fos and JunB to the IFN-y promoter in these cells. The observed
bands are unlikely to be the result of nonspecific binding because PCR reactions specific
for a portion of IFN-y intron-1, located one kb downstream of the promoter, resulted in
product when the pre-precipitate DNA sample was used, but not when any antibodyprecipitated DNA samples were used.
It should be noted that there are likely other positions where API and CREB/ATF
family members bind to the IFN-y promoter within 500 bp of the proximal AP1
CREB/ATF binding site. It is therefore probable that some of the factors that were
shown by this assay to bind to the promoter in either the TH1 cell line or the TH2 cell line
are bound at locations other than the -53 position. However, these results correlate well
with the in vitro EMSA results, and suggest that the TH1 cell line's unmethylated IFN-y
promoter is bound by CREB, ATF2 and c-Jun, whereas the TH2 cell line's methylated
IFN-y promoter is not.
To further investigate the differences in factor binding of the IFN-y promoter
between TH1 and TH2 cells, we assayed for c-Jun, CREB, ATF1/CREB, and ATF2 in
nuclear extracts of AE7 and DO10cells by Western blotting. As shown in Figure 20, the
relative levels of c-Jun and ATF1 were similar in TH1 and TH2 nuclear extracts, whereas
the relative levels of ATF2 and CREB were significantly higher in the TH1 nuclear
extracts than in the TH2 nuclear extracts. Thus, nuclear protein levels, in addition to
methylation status of the IFN-y promoter, may contribute to the observed differences in
factor binding between the THI1and TH2 cells.
Methylation of the -53 CpG inhibits the IFN-y promoter activity
While the above studies examined the effect of CpG methylation on transcription
factor binding at the IFN-y promoter, they did not address how the observed methylation
changes affected promoter function. In order to examine this issue, we performed a
series of transient transfection assays where we were able to test the ability of the IFN-y
proximal promoter to drive the expression of a luciferase reporter while methylated at
various CpGs.
99.
D10
(T H2)
AE7
(T H 1)
c-Jun
CREB
CREB
ATF1
1__
"_Al~' ::
1._
-_----'
.-
-
....
,
....
,
..
,'
~-
ATF2
Figure 20. c-Jun, CREB, ATFl, and ATF2 are present in both AE7 and 010 cells.
Western blots demonstrating relative nuclear protein levels of c-Jun, CREB, ATFl, and
ATF2 in AE7 and 010 cells.
100
For these studies, we constructed a vector, referred to as y-luc, in which the firefly
luciferase gene was under the control of the proximal 250 bp of the IFN-y promoter
(Figure 21). This construct was co-transfected with a control vector that contained CMV
driven ranilla luciferase into either the TH1 cell line AE7, which, as described above,
expressed IFN-y following PMA plus ionomycin stimulation (Figure 5), or the TH2 cell
line DIO, which did not express IFN-y following stimulation. Transfection was done by
electroporation using the Amaxa Nucleofector device.
The level of IFN-y promoter driven luciferase expression for the two cell lines
was determined both before and after PMA and ionomycin stimulation. Compared to a
vector without the IFN-y promoter (pGL3) (Figure 21), significant luciferase activities
were observed when the y-luc vector was transiently transfected into both the AE7 and
the DIO cell lines, even without PMA and ionomycin stimulation (p < 0.001), and that
luciferase expression in both cell lines increased significantly stimulation (p < 0.001)
(Figure 22). These results suggest that the 250 bp IFN-y promoter is sufficient to drive
the reporter gene expression. Furthermore, the presence of luciferase activities in the TH2
cell line is consistent with the presence of transcription factors known to be involved in
IFN-y transcription in DIO cells (Figure 20) (Yano et al., 2003; Zhang et al., 1998). The
ability of the proximal 250 bp of the IFN-y promoter to drive luciferase expression in
DO10cells, which do not express IFN-y, suggests that there is a suppressive mechanism
that is active in the endogenous promoter and that is not present in the transfected
promoter. Nevertheless, the 250 bp IFN-y promoter is significantly more active in AE7
cells than in DO10cells, both with and without PMA plus ionomycin stimulation (p < 0.02
for both).
In order to test the importance of the proximal API binding site in promoter
function, we generated a construct identical to y-luc except that this site was deleted from
the proximal promoter (Figure 21). When this construct was transfected into AE7 cells
that were then stimulated with PMA and ionomycin, the resulting luciferase expression
was still significantly higher that of pGL3 (p < 0.01), but was approximately 18 fold
lower than when a construct containing the full promoter was used (Figure 23).
To test the effect of full construct methylation on promoter function, we
methylated y--luc using mSss I (CpG methylase). When this construct was transfected
101
I
PI
l
y-luc
IFN-y promoter -250 to
+1oo00
Firefly Luciferase
I-
pGL3
I
y-luc AP1
site deletion
IFN-y pro. -250 to -57 -47 to +100
Firefly Luciferase
_
I
Firefly Luciferase
Figure 21. Schematic representation of the vectors used in the transient transfection
assays.
Schematic representations of y-luc, pGL3 basic, and the y-luc vector containing a deletion
of the proximal API binding site.
102
25
20
y-Iuc Expression is Inducible
T 1
:::>
...J
a: 15
"0
Q)
.~
co
E
~ 10
o
Z
5
o
P/I
Vector
+
pGL3
+
+
y-Iuc
pGL3
+
y-Iuc
Figure 22. The 250 bp IFN-y proximal promoter can drive luciferase expression in
both AE7 and DIO cells.
Data shown are firefly luciferase activities normalized to renilla luciferase activities
y-Iuc and pGL3 vectors transfected into AE7 or DID cells with or without PMA and
ionomycin stimulation.
The error bars indicate standard deviations from triplicate
samples in each experiment.
Representative
data from one of two independent
experiments are shown. P values were calculated by the student t-test
for
103
20.0
18.0
16.0
::>
14.0
-.J
a: 12.0
"'0
Q)
.~ 10.0
a:s
E
~
8.0
0
z
6.0
4.0
2.0
0.0
pGL3
y-Iuc
AP1 site
Deletion
y-Iuc
mSss I
Treated
y-Iuc
Figure 23. The proximal API CR EB/ ATF binding site is required
for full IFN-y
proximal promoter function.
Data shown are firefly luciferase activities normalized to renilla luciferase activities for
pGL3, y-Iuc, y-Iuc treated with mSss. I, or y-Iuc with the proximal API CREB/ATF
binding site deleted transfected into AE7 cells with 2 hours of PMA and ionomycin
stimulation. The error bars indicate standard deviations from triplicate samples in each
experiment. Representative data from one of two to three independent experiments are
shown. P values were calculated by the student t-test.
104
into PMA and onomycin stimulated T cells the resulting luciferase expression was
approximately 10 fold less than when an unmethylated construct was used (Figure 23).
This demonstrates that complete construct methylation is sufficient to inhibit promoter
function, but leaves uncertain the importance of the methylation of individual CpGs.
In order to address the effect of methylation of specific CpGs on IFN-y promoter
function, we modified the y-luc vector so that each of the promoter's six CpGs was
individually methylated or the three CpGs at -53, -45, and -34 positions were methylated
simultaneously by a PCR-mediated mutagenesis.
Using a protocol adapted from
Martinowich et al, we PCR amplified the entire y-luc construct using self-complementary
primers that contained site-specific methylation at the desired CpGs. The template
plasmid was then digested using Dpn. I, and the resulting PCR product was column
purified and concentrated. As controls, we generated unmethylated y-luc vector by the
same PCR technique.
These vectors were co-transfected into AE7 cells with a renilla luciferase plasmid
standard. The transfected cells were stimulated with PMA and ionomycin and assayed
for luciferase activities. The luciferase activities (normalized to the renilla luciferase
activities) were virtually the same in cells transfected with the unmethylated vector or
vector methylated at the -205, -190, or-170
position (Figure 24).
In contrast,
methylation at the -53 site alone reduced the luciferase activity significantly (p < 0.001),
to a level that was approximately 33% of that produced by the unmethylated PCRgenerated construct. Methylation at the -45 or -34 sites also reduced luciferase activity,
although by a lesser, but still significant, amount (p = 0.003 and p = 0.04, respectively).
However, methylation of the -53 site along with the -45 and -34 sites in the same vector
did not reduce luciferase activities more than methylation of the -53 site alone. These
results show that the methylation of the -53 CpG alone has the most pronounced effect
on IFN-y promoter activity.
105
10
8
2
-- + +- - -- -- -- --- -- -- +- +- -- -- +- - - - - + +- ++
------
o
+
Methylated CpG
-
-205
-190
-170
-53
IFN-y
Promoter
Position
-45
-34
Unmethylated CpG
Figure 24. Inhibition of IFN-y promoter activity by site-specific methylation.
The y-luc vector with the indicated methylation(s) and the pGL3 vector were transfected
with the renilla luciferase vector into AE7 cells. Six hours later, the transfected cells
were stimulated with PMA and i onomycin for two hours and luciferase activities were
measured. Data shown are firefly luciferase activities normalized to renilla luciferase
activities. (+) indicate a methylated CpG at the indicated position, while (-) indicate an
un methylated CpG at the indicated position. The error bars indicate standard deviations
from triplicate samples in each experiment.
Representative data from one of three to five
independent experiments are shown. P values were calculated by the student t-test.
106
Discussion
Dynamic relationships between methylation state and transcription at the IFN-y
locus
The methylation status of the IFN-y promoter in naive CD4 T cells has been
reported both as hypermethylated (Katamura et al., 1998; Melvin et al., 1995; White et
al., 2002) and hypomethylated (Winders et al., 2004). Furthermore, these studies only
looked at one or two CpGs within the proximal promoter by semi-quantitative southern
blotting, and made the assumption that those CpGs were representative of all the CpGs
within this region. Our study, on the other hand, provides quantitative analyses of all six
CpGs in the proximal promoter, and has unequivocally demonstrated that the IFN-y
promoter is hypomethylated in naive CD4+ T cells. Furthermore, our results show that
the IFN-y promoter already hypomethylated in precursor thymocytes.
The
hypomethylation of the promoter during precursor T cell development is contrasted by
the hypermethylation of the transcribed region of the IFN-y locus, indicating the sitespecificity of the process that established this methylation phenotype. In addition, our
analyses of the methylation status in the IFN-y locus during the course of T helper cell
differentiation reveal that the IFN-y promoter becomes de novo methylated during TH2
cell polarization, whereas the transcribed region becomes demethylated during the TH
cell polarization and memory CD8 T cell development.
Together, these results
demonstrate dynamic changes in the methylation of the IFN-y locus during T cell, and
especially T helper cell, development.
Based on our findings and previous observations, it is clear that there are three
different the IFN-y locus methylation states that correlate with the cell's potential and/or
history of IFN-y transcription. The first is the hypermethylated (or repressed) state of the
IFN-y locus. Cells with this phenotype, including kidney cells, heart cells, B cells, and
long-term TH2 cell lines, have both promoter and transcribed region hypermethylated
(Figures 2-2 and 2-3), and are programmed not to express IFN-y. The heritable nature of
)NA methylation allows this phenotype to be passed on following cell division to
daughter cells, preventing them from expressing IFN-y as well. This methylation state is
correlated with histone hypoacetylation, and promoter nuclear localization in
107
heterochromatic regions of the nucleus (Avni et al., 2002; Eivazova and Aune, 2004;
Grogan et al., 2001). Thus, the permanent repression of the IFN-y locus in these cells is
likely mediated by chromatin changes, to which CpG methylation is thought to contribute
through the recruitment of methyl-CpG binding proteins and histone deacetylases
(Richards and Elgin, 2002).
Given that the cytokine genes share many activating
transcription factors, it is also possible that methylation of the IFN-y promoter prevents
inappropriate transcription of IFN-y when other cytokines are induced in the same cells,
such as in T,, 2 cells (see below).
The second methylation phenotype we saw at the IFN-y promoter is the "ready"
(or open) state, found in cells such as naive CD4+ and CD8+ T cells and thymocytes,
which have the potential to express IFN-y, but have yet to transcribe the gene for
extended periods of time. In these cells, the IFN-y promoter is hypomethylated, but the
transcribed region is hypermethylated (Figures 5 and 6). These observations reaffirm the
notion that the promoter demethylation is not sufficient for IFN-y transcription. It is
likely that cells in this state, such as naive CD4+ T cells, are able to rely on regulation of
transcription factors to prevent improper IFN-y transcription, while at the same time
maintain the ability for rapid transcriptional induction (see below). As the histones at the
IFN-y promoter are hypoacetylated in naive helper T cells (Avni et al., 2002), it is likely
that there are also other epigenetic mechanisms for transcriptional suppression at work in
these cells.
The third methylation
phenotype found at the IFN-y promoter
is the
hypomethylated (or active) state of the IFN-y locus. Cells in this state, such as memory
CD8+ T cells, long-term polarized THi cells, and NK cells, have undergone extensive
IFN-y transcription and have a completely hypomethylated promoter and transcribed
region (Figure2 3, 7 and 11). Even when cells in this state are not expressing IFN-y
protein, they often express a low level of IFN-y transcript (Grayson et al., 2001; Stetson
et al., 2003), and may regulate IFN-y production at a post-transcriptional
level (Malter,
1998; Nagy et al., 1994). This state enables cells, such as memory CD8+T cells and THi
cells, to rapidly respond to stimulation with production of IFN-y without waiting for
transcription to occur.
It has been demonstrated that cells with this methylation
phenotype often also have histone hyperacetylation at the IFN-y promoter (Avni et al.,
108
2002), which may further contribute to the accessibility of the locus. It is also tempting
to speculate that this active state of the IFN-y locus contributes to the T-bet independent
IFN-y expression acquired by TH cells once they have been polarized for a long period
of time (Martins et al., 2005; Mullen et al., 2002).
The kinetics of methylation change during helper T cell polarization vary depending
upon CpG location
During T
cell polarization, the CpG located at the +16 position becomes
hypomethylated faster than the CpG located at the +96 position. The existence of cells in
which the +16 position CpG is methylated but the +96 position CpG is demethylated
(Figure 9) suggests that this difference in demethylation kinetics is the result of the CpG
located at the + 16 position being more susceptible to demethylation than the CpG located
at the +96 position, rather than a directional demethylation mechanism.
It is also
important to note that this demethylation process is relatively slow, suggesting passive
demethylation taking place during T cell proliferation rather than an active demethylation
process. Furthermore, IFN-y transcription occurs well before most of the T cells have
undergone demethylation, indicating that transcribed region demethylation is not required
for IFN-y expression.
The demethylation of the IFN-y promoter proximal transcribed region is similar to
the demethylation event that occurs at the IL-4 promoter (Lee et al., 2002). In these
instances, the demethylation
occurs slowly enough to be attributed to a passive
demethylation process, occurs faster at the more 5' CpGs, and occurs after transcription
is initiated. These similarities suggest that a similar demethylation mechanism may be
responsible for both events.
While the demethylation event that occurs at the IL-4
promoter is not required for IL-4 expression, cells that are hypomethylated at this
location express higher levels of IL-4 than IL-4 promoter hypermethylated cells (Lee et
al., 2002). It would be interesting to investigate whether a similar correlation exists
between IFN-y expression level and the methylation status of the promoter proximal
transcribed region CpGs.
It is also notable that de novo methylation does not occur with the same kinetics at
all six CpGs within the IFN-y proximal promoter during TH2 cell polarization (Figure 15).
109
The methylation does not spread in a simple pattern, e.g., from 5' to 3' or from 3' to 5',
but rather seems to pick and choose CpGs, with some CpGs becoming methylated faster
than others. While the -53 CpG becomes methylated faster than the other CpGs within
the proximal promoter, there are instances where other CpGs are methylated but the -53
CpG is not methylated. This suggests that the methylation machinery may be targeted to
particular CpGs more readily than others, but implies that there is not a specific order to
the de novo methylation events. While the mechanism behind this de novo methylation is
unknown, it is interesting to note that the -53 CpG remains unmethylated when the T
cells are activated in the absence of polarizing conditions. This suggests that the signals
that induce other aspects of TH2 cell development also induce the methylation of the -53
CpG, rather than methylation occurring as a result of T cell activation in the absence of
TH1 signals.
Methylation of the -53 CpG inhibits transcription factor binding to the IFN-y
promoter
The rapid methylation at the conserved -53 CpG during the course of TH2 cell
polarization may affect factor binding to the CpG-containing proximal AP1 site. In
support of this hypothesis, our EMSA studies demonstrate that two complexes from the
nuclear extract of the TH1 cell line AE7 are capable of binding to the proximal AP1 site
of the IFN-y promoter in vitro (Figure 16). One complex contains CREB and the other
ATF2 and c-Jun (Figure 18), consistent with previous reports of AP1 and CREB/ATF
binding at this site (Cippitelli et al., 1995; Penix et al., 1996; Zhang et al., 1998).
Importantly, we show that methylation of the -53 CpG alone, but not the -45 or -34
CpGs, was sufficient to reduce the levels of factor binding by greater than 10 fold (Figure
17). These findings are consistent with previous observations showing that methylation of
the -53 CpG results in a change in factor binding to the site, although the early study did
not identify the specific transcription factors involved (Young et al., 1994). Our findings
are also consistent with the observation that CpG methylation can inhibit the binding of
some AP1 family members to certain DNA sequences, such as in the TrkA and
neurotensin/neuromedin
N genes (Dong et al., 2000; Fujimoto et al., 2005).
Complementing the in vitro binding results, we found that the same factors are bound to
110
the hypomethylated
IFN-y promoter in a TH I cell line but not the hypermethylated
IFN-y
promoter in a TH2 cell line (Figure 19). While the ChIP assay is unable to demonstrate
exactly where in the promoter the factors are bound, the ChIP data correlates with the
EMSA data. Together, these findings suggest that methylation of the -53 CpG inhibits
transcription factor binding to the IFN-y promoter both in vitro and in vivo.
Methylation of the -53 CpG inhibits IFN-y promoter activity
The effect of the -53 CpG methylation on the IFN-y promoter activity was
directly examined by reporter assays in a TH 1 cell line, in which the IFN-y is transcribed
(Figure 4). We showed that the 250 bp IFN-y promoter is capable of driving a significant
reporter gene expression in both a TH 1 cell line and a TH2 cell line, especially after PMA
plus ionomycin stimulation (Figure 22). The existence of promoter activity in a TH2 cell
line is consistent with the presence of activating transcription factors in the nucleus of
these cells (Figure 20). The fact that the IFN-y proximal promoter located in the
transcribed vector was able to drive luciferase expression, while the endogenous IFN-y
vector was not able to drive IFN-y expression, is further evidence for a suppressive
mechanism that affects the endogenous promoter more than the transfected promoter.
This is not surprising, considering that the endogenous promoter is nearly completely
hypermethylated in DO10cells (Figure 3), and is probably packaged in hypoacetylated
histones and localized in a heterochromatic region of the nucleus. Furthermore, there
may be additional suppressive cis-elements present in the larger endogenous locus that
are not in the proximal promoter.
It is clear that the proximal API site is critical for the promoter activity because
its deletion from the vector results in an almost complete abolition of promoter activity
(Figure 23). Importantly, methylation of the -53 CpG alone, but not methylation of the
-205, the -190 or the -170 CpGs in the reporter vector results in a significant reduction of
luciferase activity (Figure 24). This suggests that methylation of the -53 CpG initiates a
location specific inhibitory mechanism that represses IFN-y proximal promoter function.
Surprisingly, methylation of the proximal promoter at the -45 or the -34 CpG
also results in a significant reduction of luciferase activity, though methylation at these
sites do not affect the observed factor binding to the proximal AP1 site in vitro. It is
111
possible that the inhibition of transcription by methylation of the -45 or -34 CpG in vivo
is the result of recruitment of 5-methylcytosine binding proteins that interfere with
binding of activating transcription factors to the proximal API binding site. This
possibility is strengthened because mice deficient in MBD2 show disregulation of IFN-y
expression (Hutchins et al., 2005), suggesting that MBD2 may play a role in the
suppression of the locus. Alternatively, it is possible that there are other factors that
contribute to induction of transcription of the reporter construct that may not be able to
exert their effect unless the -53 CpG site is also occupied. In support of this
interpretation, methylation of the -53, -45, and -34 CpGs simultaneously does not reduce
luciferase activities more than methylation of the -53 CpG alone. Together, these results
suggest a critical role of the methylation of the conserved -53 CpG in inhibiting IFN-y
promoter activity. They further suggest that methylation of other CpGs in the vicinity
may help to suppress the promoter activity.
Regulation of IFN-y transcription in helper T cell development
When nuclear extract from the TH2 cell line D10 was used in the EMSA
experiments,
methylation specific factor binding was found to be similar to the THI cell
line AE7. Similarly, luciferase expression driven by the 250 bp IFN-y promoter was also
seen in DO10cells, especially following PMA plus ionomycin stimulation. As the API,
ATF and CREB proteins are important transcription factors for many genes, it is not
surprising that these factors are not specific for THI cells. The presence and similar
binding ability of these transcription factors to the IFN-y promoter in both THI and TH2
cells and the inhibition of TH2 polarization by IFN-y underscores the need to rapidly
suppress IFN-y transcription in developing TH2 cells by other mechanisms. Our findings
suggest that the rapid and site-specific methylation of the promoter likely contributes to
this regulation.
By rapidly repressing the "ready" state of the IFN-y promoter, the
methylation event helps to promote TH2 cell development.
How does promoter methylation contribute to the overall regulation of the IFN-y
transcription in CD4+ T cells? It has been shown that a dominant negative variant of the
AP1 family member c-Jun inhibits IFN-y expression (Cippitelli et al., 1995; Penix et al.,
1996), implicating that this factor may normally activate IFN-y transcription. In contrast,
112
overexpression of CREB was shown to inhibit IFN-y promoter function (Penix et al.,
1993; Zhang et al., 1998), indicating that CREB may normally suppress IFN-y
transcription. Upon activation of CD4+ T cells, regardless of polarizing conditions, CREB
levels are reduced whereas c-Jun levels remain unchanged (Zhang et al., 1998),
suggesting that the ratios of these two transcription factors may play a critical role in
IFN-y transcription.
Thus, we propose a general model that integrates the contributions of both
promoter CpG methylation and ATF2/c-Jun and CREB in IFN-y transcriptional
regulation in CD4 + T cells. In naive CD4 + T cells, the IFN-y promoter is hypomethylated
but IFN-y is not transcribed. It is possible that, in these cells, the negatively acting CREB
contributes to the suppression of the locus, perhaps by preventing the positively acting cJun from binding to a common or overlapping cis-element (Cippitelli et al., 1995; Penix
et al., 1996; Zhang et al., 1998) (Figure 25A). Upon CD4+ T cell activation, nuclear
CREB, but not c-Jun, levels are reduced (Zhang et al., 1998), this change in c-Jun to
CREB ratio may allow c-Jun to interact with the hypomethylated IFN-y promoter,
contributing to IFN-y transcription in newly activated CD4+ T cells and in T.1 polarized
cells (Figure 25B). Under TH2 polarizing conditions, the rapid methylation of the -53
CpG in the IFN-y promoter directly inhibits c-Jun/ATF2 binding and therefore prevents
inappropriate IFN-y transcription during the course of TH2 cell polarization (Figure 25C).
As the TH2 polarization progresses, the remaining CpGs in the IFN-y promoter gradually
become hypermethylated, contributing to further transcription repression, chromatin
remodeling and the permanent suppression of the locus (Figure 25D).
In conclusion, our results show that the IFN-y promoter is hypomethylated in
naive CD4+ T cells and becomes progressively methylated during TH2 cell polarization.
The rapid methylation of the -53 CpG probably prevents IFN-y transcription immediately
by directly inhibiting transcription factor binding to the promoter. As the -53 CpG is
conserved in the IFN-y promoter among mouse, rat, dog, chimpanzee and human, the
rapid methylation of the -53 CpG is likely a general mechanism for repressing IFN-y
transcription during TH2 cell polarization. The more extensive methylation of the IFN-y
promoter in polarized TH2 cells probably contributes to the permanent silencing of the
locus by mediating chromatin remodeling.
113
A
NaYve CD4 T cells
??
B
CR
- U
AB
-53
Recently polarized
TH1 cells
-53
C Recentlypolarized,,
TH2 cells
-??53
-53
D
Long-term polarized TH2 cells
-53
T
Methylated CpG
?
Unmethylated CpG
Figure 25. A schematic diagram depicting the relationship among CpG methylation,
AP1 and CREB/ATF transcription factor binding, and IFN-y transcription.
(A) In naive CD4 T cells, the high level of the IFN-y inhibitory factor CREB relative to
an activating complex containing c-Jun and ATF2 contributes to suppression of the
hypomethylated IFN-y promoter. (B) Activation under TH 1 polarizing conditions leads to
a reduction in the nuclear levels of CREB, enabling the c-Jun/ATF2 complex to bind to
the promoter and contribute to IFN-y transcription. (C) Activation under TH2 polarizing
conditions leads to a reduction in the nuclear levels of CREB, but rapid methylation of
the -53 CpG prevents c-Jun/ATF2 from binding to the promoter, and thus contributes to
the suppression of IFN-y transcription. (D) In long-term polarized TH2 cells, the
remaining CpGs in the IFN-y pro moter are methylated, contributing to chromatin
remodeling and t he permanent suppression of t he locus. For simplicity, only the
methylation status of the -53 CpG is considered.
114
Materials and Methods
Cell Preparation and Culture
Primary naive CD4* T cell populations
(CD4+, CD45RB h gI h , CD62L h g h, CD44ow)
and B-cell populations (B220+) were purified by flow cytometry from lymph nodes and
spleens of Balb/c mice. CD4 and CD8 double positive thymocytes, CD8* thymocytes and
CD4+ thymocytes were purified from thymuses of C57B16 mice by flow cytometry. CD4
and CD8 double negative thymocytes were sorted from thymuses of RAGl-deficient
C57BL6 mice, and Natural Killer cells (Pan-NK+, NK-1. 1+)were sorted from spleens of
RAG Il-deficientC57BL6 mice. Naive CD8* T cells (CD8+, CD44'°W) were sorted from
spleens and lymph nodes of Balb/c mice. All antibodies used in flow cytometry were
obtained from BD Pharmingen, San Diego, CA. Purity of all cell populations was greater
than 95%.
Sorted naive CD4* T cells were stimulated at a density of
x 106/ml with 5 [ig/ml
plate-bound anti-CD3 and 5 tg/ml plate-bound anti-CD28 in the presence of polarizing
cytokines. THI polarized cultures contained 5 ng/ml recombinant IL-12 and 10 gg/ml
neutralizing anti-IL-4 antibody (BD Pharmingen), while TH2 polarized cultures contained
25 ng/ml recombinant IL-4 and 10 [xg/ml neutralizing anti-IFN-y antibody (BD
Pharmingen). THO cells contained both 10 [tg/ml anti-IL-4 antibody and 10 gg/ml anti
IFN-y antibody, but no polarizing cytokines. After 24 hours, 40 U/ml IL-2 was added to
all cultures. Cultures were washed and fresh cytokines were added every 3 days and
restimulated under the above conditions every 6-7 days.
Effector CD8 T cells were generated by stimulating sorted naive CD8+ T cells at
a density of
x 106 /ml with 5
g/ml plate-bound anti-CD3 and 5
g/ml plate-bound anti-
CD28 for three days. Memory phenotype CD8 T cells were generated by sorting naive
CD8* T cells from spleens and lymph nodes of RAG 1-deficient C57BL6 mice expressing
a 2C transgenic TCR. These cells were adoptively transferred intravenously into RAG 1deficient C57BL/6 mice, which were then immunized with SIYRYYGL peptide in CFA.
Approximately 6 months later, CD8+CD44himemory 2C T cells were sorted from spleens
and lymph nodes of these mice.
115
Cytokine Recapture Staining
Cytokine recapture staining was done using either the Mouse IFN-y Secretion
Assay kit or the Mouse IL-4 Secretion Assay kit (both from Miltenyi Biotec, Auburn,
CA) according to the manufacturer's instructions. Briefly, T cell populations were either
restimulated with 5 ng/ml PMA and 100 ng/ml ionomycin in 10% FCS RPMI media or
cultured in 10% FCS RPMI media without PMA or ionomycin for 2 or 3 hours. Cells
were washed and stained in 100 [d 5% FCS RPMI containing 10 ¢1 of the catch antibody
in a 50 ml conical tube on ice. After 5 minutes, 30 ml of warm 5% FCS RPMI was added
to the tubes, and cells were incubated with constant rotation at 37 degrees for 45 minutes.
Cells were washed with cold 1.5 mM EDTA 0.5% BSA PBS buffer and stained in a 100
tl volume with 10 la detection antibody and 1 1tlAPC conjugated anti-IL-4 antibody.
Cell were cultured on ice for 15 minutes, washed, and analyzed by FACS.
Bisulfate Conversion and Methylation Analysis
Figure 26 is a schematic for bisulfite sequencing methylation analysis.
Methylation analysis was done by bisulfite conversion of genomic DNA using the
CpGenome Universal DNA Modification Kit (Chemicon International, Temecula, CA)
according to the manufacturer's
amplified
using
instructions.
the
The IFN-y promoter was then PCR
following
primers:
TAGAGAATTTTATAAGAATGGTATAGGTGGGTAT
and
Forward:
Reverse:
CCATAAAAAAAAACTACAAAACCAAAATACAATA. The initial annealing
temperature for the PCR reaction was 56 degrees.
The annealing temperature was
reduced at a rate of 0.5 degrees per cycle until an annealing temperature of 47 degrees
was reached. The PCR reaction was run at this annealing temperature for 25 more
cycles, followed by a 5-minute elongation step. The PCR product was cloned using the
TOPO TA Cloning kit (Invitrogen, Carlsbad, CA), according to the manufacturer's
instructions, and clones were sequenced using the included primers.
Electrophoretic Mobility Shift Assay
Nuclear extracts of AE7 and D10 T cell lines were generated as previously
described (Schreiber et al., 1989) except that a Complete Mini protease inhibitor cocktail
116
Bisulfite Conversion Assay
Unmethylated CpGs
Methylated CpGs
CH 3
ATACGAATCCCGA
OH 3
CH 3
OH 3
ATAC(GAATCC!CGA
I
Sodium bisulfite treatment deaminates unmethylated cytosine
(converting them to uracil) but leaves methylated cytosine intact
CH 3
ATAUGAATUUUGA
CH3
ATAGAATU UGA
~~~~ATA
GAATUUGA
]
l
I
The locus of interest is amplified by PCR, which
converts the uracil bases to thymidine
ATATGAATTTTGA
ATACGAATTTCGA
The PCR product is cloned and sequenced. The ratio of
cytosine to thymidine at a CpG is used to determine the
original methylation state of that base
4
Figure 26. The bisulfite conversion assay.
A flow chart illustrating the steps of the sodium bisulfite sequencing method for
methylation analysis. Genomic DNA is first treated with sodium bisulfite, followed by
deamination. This converts unmethylated cytosine into uracil, but methylated cytosine
are resistant to conversion. This is followed by PCR amplification of the locus of
interest, during which uracil is converted to thymidine. The PCR product is cloned and
sequenced. The m ethylation percentage is determined by comparing the amount of
cytosine to the amount of thymidine at a particular position.
117
tablet (Roch, Mannheim, Germany) was included in buffers A and C. DNA binding
reactions were performed for 30 minutes at room temperature as previously described
(Shen and Stavnezer, 2001).
For competition experiments, a 1, 10 or 100 fold molar
excess of the specified unlabeled competitor oligonucleotide were added to the reaction.
For Ab supershift experiments, 200 ng of the indicated Ab were added to the reactions.
The binding reactions were electrophoresed on a 5% native polyacrylamide gel followed
by autoradiography.
The following oligonucleotide sequence correlating to the -62
through -32 position of the IFN-y promoter was used with the three included CpGs either
methylated or unmethylated: GTGAAAATACGTAATCCCGAGGAGCCTTCGA.
NFAT
consensus
oligonucleotide
used
TATGAAACAAATTTTCCTCTTTGGGCG.
had
the
following
The
sequence:
The following Abs were used for the
supershift assay: Fos B (sc-48) Jun B (sc-46) c-Fos (sc-52), c-Jun (sc-45), ATFI/CREB
(sc-270), ATF2 (sc-6233) from Santa Cruz Biotechnology (Santa Cruz, CA) and CREB
(35-0900) from Zymed (San Francisco, CA).
Western Blotting
Nuclear extracts of AE7 and D10 cell lines were generated as described above.
50 tg/lane of the nuclear extracts were fractionated by SDS-PAGE and transferred to a
nitrocellulose membrane.
The membranes were probed with the specific primary
antibodies described above. Membranes were washed and incubated with horseradish
peroxidase linked antibodies against either mouse IgG or rabbit IgG (Amersham,
Piscataway, NJ). Blots were visualized using Western Lightning chemiluminescence
reagent (PerkinElmer, Boston, MA) according to the manufacturer's instructions.
Chromatin Immunoprecipitation
The Chromatin Immunoprecipitation
assay was done using the Chromatin
Immunoprecipitation (ChIP) Assay Kit from Upstate (Lake Placid, NY) according to
manufacturor's instructions. Briefly, stimulated AE7 and DlO cells were treated with 1%
formaldehyde for 10 minutes at 37 degrees.
Cells were lysed and the lysate was
sonicated with 4 20-second pulses, generating DNA fragments with sizes ranging from
200 to 600 bp. The antibodies used for the EMSA experiments and a Protein A bound
118
Agarose slurry were used to precipitate the targeted complexes.
The formaldehyde
cross-linking was reversed by incubating the complexes in 200 mM NaCl at 65 degrees
for 4 hours. Presence of the IFN-y promoter sequences was determined using the
following PCR primers: Forward: GCTGTCTCATCGTCAGAGAGCCCA and Reverse:
TGATCGAAGGCTCCTCGGGATTACG. The following primers were used to amplify
a region of intron 1: Forward: CAGTAACAGTGTTTGGCTACATGC
and Reverse:
ACCTGCCCTGAAAATATCTATCA. PCRs were run with an annealing temperature of
54 degrees.
Vector Construction and Methylation
The sequences between -250 and +100 relative to the IFN-y transcriptional start
site was amplified by PCR and inserted into the Sac I and Xho I sites of the pGL3 basic
reporter vector (Promega). The resulting vector is referred to as y-luc. The Quickchange
Site-Directed Mutagenesis Kit (Stratagene) was used to delete the proximal API site (-47
through -57) in the y-luc vector.
Methylation of the above construct was either done using CpG Methylase (mSss.
1) (New England Biolabs, Beverly, MA), or using the Quickchange Site-Directed
Mutagenesis Kit and a protocol adapted from Martinowich et al. (Martinowich et al.,
2003). A schematic for this protocol can be found in Figure 27. Briefly, the y-luc vector
was methylated specifically at the -205. -190, -170, -53, -45, or -34 CpG using sitespecific methylated oligonucleotides purchased from IDT (Coralville, IA). The following
primers and their reverse complements,
unmethylated
containing either methylated CpGs or
CpGs, were used for this experiment:
CpGs -53, -45 and -34:
GTGAAAATACGTAATCCCGAGGAGCCTTCGATCAGGTATAAAAC
-190:
-205 and
GGGCACAGCGGGGCTGTCTCATCGTCAGAGAGCCCAAGG
CGTCAGAG(AGCCCAAGGAGTCGAAAGGAAACTCTAAC.
-170:
PCR was done
according to the manufacture's protocol with 100 ng/reaction of the y-luc vector used as a
template. Following the PCR, the reaction was treated with Dpn I for 2 hours to remove
the template DNA. The product was purified using the Qia-Quick method from Qiagen,
re-annealed, and precipitated. DNA concentration was estimated by gel electrophoresis.
119
IV%,IIIL,
I I PULI lyl-L--
-.I'.--,Pi 11I Im 0
PCR Amplification
I
IT
I
Dpn. I Digestion
Template Plasmid
Template Plasmid
0
I
Figure 27. Site-specific methylation of a plasmid
A diagram illustrating the method used to generate site-specific methylated vectors. The
entire template vector is amplified by PCR using site-specific methylated primers that are
reverse complements of each other. The product of this reaction is nicked, site specific
methylated vectors. The PCR product is then digested with Dpn. I, which digests the
template, but not the product. The digest is then column purified, re-annealed, and
concentrated by ethanol precipitation.
120
Cell Transfection and Luciferase Detection
AE7 and DIO cells were transfected using the Amaxa Nucleofector Device and
the Amaxa Mouse T Cell Nucleofector Kit (Amaxa, Koelin, Germany) following the
manufacturer's protocol. Briefly, 2 x 106AE7 or DIO cells per reaction were harvested
and washed in 0.5% BSA in PBS. Cells were resuspended in 100 l nucleofector
solution, to which 2 g of the test vector and 100 ng of a control CMV promoter driven
renilla luciferase construct were added. After electroporation, the cells were cultured in
1.5 ml of the Mouse T Cell Nucleofector Medium for 6 hours at 37 degrees. The cells
were then restimulated with 5 ng/ml PMA and 100 ng/ml onomycin for two hours,
following which cells were harvested and lysed. Luciferase levels were determined using
the Dual Reporter Luciferase Assay System (Promega, Madison, WI) according to the
manufacture's instructions.
Acknowledgements
I thank Dr. Laurie Glimcher for the AE7 and DIO cell lines; Dr. Qing Ge for
assistance in the generation of memory CD8+T cells; and Dr. Ching-Hung Shen for help
with the EMSA assays.
121
References
Afkarian, M., Sedy, J. R., Yang, J., Jacobson, N. G., Cereb, N., Yang, S. Y., Murphy, T.
L., and Murphy, K. M. (2002). T-bet is a STAT 1-induced regulator of IL-12R expression
in naive CD4+ T cells. Nat Immunol 3,549-557.
Agarwal, S., and Rao, A. (1998). Modulation of chromatin structure regulates cytokine
gene expression during T cell differentiation. Immunity 9, 765-775.
Ansel, K. M., Lee, D. U., and Rao, A. (2003). An epigenetic view of helper T cell
differentiation. Nat Immunol 4, 616-623.
Aune, T. M., Penix, L. A., Rincon, M. R., and Flavell, R. A. (1997). Differential
transcription directed by discrete gamma interferon promoter elements in naive and
memory (effector) CD4 T cells and CD8 T cells. Mol Cell Biol 17, 199-208.
Avni, O., Lee, D., Macian, F., Szabo, S. J., Glimcher, L. H., and Rao, A. (2002). T(H)
cell differentiation is accompanied by dynamic changes in histone acetylation of cytokine
genes. Nat Immunol 3, 643-651.
Chitnis, T., and Khoury, S. J. (2003). Cytokine shifts and tolerance in experimental
autoimmune encephalomyelitis. Immunol Res 28, 223-239.
Cippitelli, M., Sica, A., Viggiano, V., Ye, J., Ghosh, P., Birrer, M. J., and Young, H. A.
(1995). Negative transcriptional regulation of the interferon-gamma promoter by
glucocorticoids and dominant negative mutants of c-Jun. J Biol Chem 270, 12548-12556.
Dong, Z., Wang, X., and Evers, B. M. (2000). Site-specific DNA methylation contributes
to neurotensin/neuromedin N expression in colon cancers. Am J Physiol Gastrointest
Liver Physiol 279, G1 139-1147.
Eivazova, E. R., and Aune, T. M. (2004). Dynamic alterations in the conformation of the
Ifng gene region during T helper cell differentiation. Proc Natl Acad Sci U S A 101, 251-
256.
Fujimoto, M., Kitazawa, R., Maeda, S., and Kitazawa, S. (2005). Methylation adjacent to
negatively regulating AP-1 site reactivates TrkA gene expression during cancer
progression. Oncogene 24, 5108-5118.
Grayson, J. M., Murali-Krishna,
K., Altman, J. D., and Ahmed, R. (2001). Gene
expression in antigen-specific CD8+ T cells during viral infection. J Immunol 166, 795799.
Grogan, J. L., and Locksley, R. M. (2002). T helper cell differentiation:
on again, off
again. Curr Opin Immunol 14, 366-372.
122
Grogan, J. L., Mohrs, M., Harmon, B., Lacy, D. A., Sedat, J. W., and Locksley, R. M.
(2001). Early transcription and silencing of cytokine genes underlie polarization of T
helper cell subsets. Immunity 14, 205-215.
Hsieh, C. S., Heimberger, A. B., Gold, J. S., O'Garra, A., and Murphy, K. M. (1992).
Differential regulation of T helper phenotype development by interleukins 4 and 10 in an
alpha beta T-cell-receptor transgenic system. Proc Natl Acad Sci U S A 89, 6065-6069.
Hutchins, A. S., Artis, D., Hendrich, B. D., Bird, A. P., Scott, P., and Reiner, S. L.
(2005). Cutting edge: a critical role for gene silencing in preventing excessive type 1
immunity. J Immunol 175, 5606-5610.
Hwang, E. S., Szabo, S. J., Schwartzberg,
P. L., and Glimcher, L. H. (2005). T helper cell
fate specified by kinase-mediated interaction of T-bet with GATA-3. Science 307, 430433.
Jankovic, D., Liu, Z., and Gause, W. C. (2001). Th - and Th2-cell commitment during
infectious disease: asymmetry in divergent pathways. Trends Immunol 22, 450-457.
Jeltsch, A. (2002). Beyond Watson and Crick: DNA methylation and molecular
enzymology of DNA methyltransferases. Chembiochem 3, 274-293.
Jenuwein, T., and Allis, C. D. (2001). Translating the histone code. Science 293, 10741080.
Jones, P. L., Veenstra, G. J., Wade, P. A., Vermaak, D., Kass, S. U., Landsberger, N.,
Strouboulis, J., and Wolffe, A. P. (1998). Methylated DNA and MeCP2 recruit histone
deacetylase to repress transcription. Nat Genet 19, 187-191.
Katamura, K., Fukui, T., Kiyomasu, T., Iio, J., Tai, G., Ueno, H., Heike, T., Mayumi, M.,
and Furusho, K. (1998). IL-4 and prostaglandin E2 inhibit hypomethylation of the 5'
regulatory region of IFN-gamma gene during differentiation of naive CD4+ T cells. Mol
Immunol 35, 39-45.
Lee, D. U., Agarwal, S., and Rao, A. (2002). Th2 lineage commitment and efficient IL-4
production involves extended demethylation of the IL-4 gene. Immunity 16, 649-660.
Lee, H. J., Takemoto, N., Kurata, H., Kamogawa, Y., Miyatake, S., O'Garra, A., and
Arai, N. (2000). GATA-3 induces T helper cell type 2 (Th2) cytokine expression and
chromatin remodeling in committed Thl cells. J Exp Med 192, 105-115.
Li, L., Young, D., Wolf, S. F., and Choi, Y. S. (1996). Interleukin-12 stimulates B cell
growth by inducing IFN-gamma. Cell Immunol 168, 133-140.
Lighvani, A. A., Frucht, D. M., Jankovic, D., Yamane, H., Aliberti, J., Hissong, B. D.,
Nguyen, B. V., Gadina, M., Sher, A., Paul, W. E., and O'Shea, J. J. (2001). T-bet is
rapidly induced by interferon-gamma in lymphoid and myeloid cells. Proc Natl Acad Sci
U S A 98, 15137-15142.
123
Malter, J. S. (1998). Posttranscriptional regulation of mRNAs important in T cell
function. Adv Immunol 68, 1-49.
Martinowich, K., Hattori, D., Wu, H., Fouse, S., He, F., Hu, Y., Fan, G., and Sun, Y. E.
(2003). DNA methylation-related chromatin remodeling in activity-dependent BDNF
gene regulation. Science 302, 890-893.
Martins, G. A., Hutchins, A. S., and Reiner, S. L. (2005). Transcriptional activators of
helper T cell fate are required for establishment but not maintenance of signature
cytokine expression. J Immunol 175, 5981-5985.
Melvin, A. J., McGurn, M. E., Bort, S. J., Gibson, C., and Lewis, D. B. (1995).
Hypomethylation of the interferon-gamma gene correlates with its expression by primary
T-lineage cells. Eur J Immunol 25, 426-430.
Mosmann, T. R., Cherwinski, H., Bond, M. W., Giedlin, M. A., and Coffman, R. L.
(1986). Two types of murine helper T cell clone. I. Definition according to profiles of
lymphokine activities and secreted proteins. J Immunol 136, 2348-2357.
Mullen, A. C., High, F. A., Hutchins, A. S., Lee, H. W., Villarino, A. V., Livingston, D.
M., Kung, A. L.., Cereb, N., Yao, T. P., Yang, S. Y., and Reiner, S. L. (2001). Role of T-
bet in commitment of TH1 cells before IL-12-dependent selection. Science 292, 19071910.
Mullen, A. C., Hutchins, A. S., High, F. A., Lee, H. W., Sykes, K. J., Chodosh, L. A., and
Reiner, S. L. (2002). Hlx is induced by and genetically interacts with T-bet to promote
heritable T(H)I gene induction. Nat Immunol 3, 652-658.
Murphy, K. M., Ouyang, W., Farrar, J. D., Yang, J., Ranganath, S., Asnagli, H.,
Afkarian, M., and Murphy, T. L. (2000). Signaling and transcription in T helper
development. Annu Rev Immunol 18,451-494.
Murphy, K. M., and Reiner, S. L. (2002). The lineage decisions of helper T cells. Nat
Rev Immunol 2, 933-944.
Nagy, E., Buhlmann, J. E., Henics, T., Waugh, M., and Rigby, W. F. (1994). Selective
modulation of IFN-gamma mRNA stability by IL-12/NKSF. Cell Immunol 159, 140-151.
Nan, X., Ng, H. H., Johnson, C. A., Laherty, C. D., Turner, B. M., Eisenman, R. N., and
Bird, A. (1998). Transcriptional repression by the methyl-CpG-binding protein MeCP2
involves a histone deacetylase complex. Nature 393, 386-389.
Pai, S. Y., Truitt, M. L., and Ho, I. C. (2004). GATA-3 deficiency abrogates the
development and maintenance of T helper type 2 cells. Proc Natl Acad Sci U S A 101,
1993-1998.
Pang, Y., Norihisa, Y., Benjamin, D., Kantor, R. R., and Young, H. A. (1992). Interferon-
gamma gene expression in human B-cell lines: induction by interleukin-2, protein kinase
124
C activators, and possible effect of hypomethylation on gene regulation. Blood 80, 724732.
Pearce, E. J., Caspar, P., Grzych, J. M., Lewis, F. A., and Sher, A. (1991).
Downregulation of Th 1 cytokine production accompanies induction of Th2 responses by
a parasitic helminth, Schistosoma mansoni. J Exp Med 173, 159-166.
Penix, L., Weaver, W. M., Pang, Y., Young, H. A., and Wilson, C. B. (1993). Two
essential regulatory elements in the human interferon gamma promoter confer activation
specific expression in T cells. J Exp Med 178, 1483-1496.
Penix, L. A., Sweetser, M. T., Weaver, W. M., Hoeffler, J. P., Kerppola, T. K., and
Wilson, C. B. (1996). The proximal regulatory element of the interferon-gamma
promoter mediates selective expression in T cells. J Biol Chem 271, 31964-31972.
Richards, E. J., and Elgin, S. C. (2002). Epigenetic codes for heterochromatin formation
and silencing: rounding up the usual suspects. Cell 108, 489-500.
Romagnani, S. (11994).Lymphokine production by human T cells in disease states. Annu
Rev Immunol 12, 227-257.
Schreiber, E., Matthias, P., Muller, M. M., and Schaffner, W. (1989). Rapid detection of
octamer binding proteins with 'mini-extracts', prepared from a small number of cells.
Nucleic Acids Res 17, 6419.
Scott, P., Pearce, E., Cheever, A. W., Coffman, R. L., and Sher, A. (1989). Role of
cytokines and CD4+ T-cell subsets in the regulation of parasite immunity and disease.
Immunol Rev 112, 161-182.
Seder, R. A., Paul, W. E., Davis, M. M., and Fazekas de St Groth, B. (1992). The
presence of interleukin 4 during in vitro priming determines the lymphokine-producing
potential of CD4+ T cells from T cell receptor transgenic mice. J Exp Med 176, 10911098.
Shen, C. H., and Stavnezer, J. (2001). Activation of the mouse Ig germline epsilon
promoter by I L-4 is dependent on AP-1 transcription factors. J Immunol 166, 411-423.
Spellberg, B., and Edwards, J. E., Jr. (2001). Type 1/Type 2 immunity in infectious
diseases. Clin Infect Dis 32, 76-102.
Stetson, D. B., Mohrs, M., Reinhardt, R. L., Baron, J. L., Wang, Z. E., Gapin, L.,
Kronenberg, M., and Locksley, R. M. (2003). Constitutive cytokine mRNAs mark natural
killer (NK) and NK T cells poised for rapid effector function. J Exp Med 198, 1069-1076.
Szabo, S. J., Kim, S. T., Costa, G. L., Zhang, X., Fathman, C. G., and Glimcher, L. H.
(2000). A novel transcription factor, T-bet, directs Thl lineage commitment. Cell 100,
655-669.
125
Szabo, S. J., Sullivan, B. M., Stemmann, C., Satoskar, A. R., Sleckman, B. P., and
Glimcher, L. H. (2002). Distinct effects of T-bet in TH I lineage commitment and IFNgamma production in CD4 and CD8 T cells. Science 295, 338-342.
Usui, T., Nishikomori, R., Kitani, A., and Strober, W. (2003). GATA-3 suppresses Thl
development by downregulation of Stat4 and not through effects on IL-I 2Rbeta2 chain or
T-bet. Immunity 18, 415-428.
White, G. P., Watt, P. M., Holt, B. J., and Holt, P. G. (2002). Differential patterns of
methylation of the IFN-gamma promoter at CpG and non-CpG sites underlie differences
in IFN-gamma gene expression between human neonatal and adult CD45RO- T cells. J
Immunol 168, 2820-2827.
Winders, B. R., Schwartz, R. H., and Bruniquel, D. (2004). A distinct region of the
murine IFN-gamma promoter is hypomethylated from early T cell development through
mature naive and Thl cell differentiation, but is hypermethylated in Th2 cells. J Immunol
173, 7377-7384.
Yano, S., Ghosh, P., Kusaba, H., Buchholz, M., and Longo, D. L. (2003). Effect of
promoter methylation on the regulation of IFN-gamma gene during in vitro
differentiation of human peripheral blood T cells into a Th2 population. J Immunol 171,
2510-2516.
Young, H. A., Ghosh, P., Ye, J., Lederer, J., Lichtman, A., Gerard, J. R., Penix, L.,
Wilson, C. B., Melvin, A. J., McGurn, M. E., and et al. (1994). Differentiation of the T
helper phenotypes by analysis of the methylation state of the IFN-gamma gene. J
Immunol 153, 3603-3610.
Zhang, D. H., Cohn, L., Ray, P., Bottomly, K., and Ray, A. (1997). Transcription factor
GATA-3 is differentially expressed in murine Thl and Th2 cells and controls Th2specific expression of the interleukin-5 gene. J Biol Chem 272, 21597-21603.
Zhang, F., Wang, D. Z., Boothby, M., Penix, L., Flavell, R. A., and Aune, T. M. (1998).
Regulation of the activity of IFN-gamma promoter elements during Th cell
differentiation. J Immunol 161, 6105-6112.
Zhu, J., Min, B., Hu-Li, J., Watson, C. J., Grinberg, A., Wang, Q., Killeen, N., Urban, J.
F., Jr., Guo, L., and Paul, W. E. (2004). Conditional deletion of Gata3 shows its essential
function in T(H)1-T(H)2 responses. Nat Immunol 5, 1157-1165.
126
Chapter 3:
The Roles of Specific Methyltransferases in the
Methylation Changes at the IFN-y Locus during TH2
Polarization
Brendan Jones and Jianzhu Chen
127
Summary
There is a strong correlation between CpG methylation and gene expression whereby
genes that are undergoing transcription are usually hypomethylated and silent genes are
usually hypermethylated. A dynamic change in methylation status occurs at the IFN-y
locus during the polarization of CD4+ T cells, with the promoter undergoing de novo
methylation during TH2 polarization,
but remaining hypomethylated
during TH
polarization. Here we show that the de novo methyltransferase Dnmt3b is not required
for this process, and that de novo methylation is reduced in Dnmt3a deficient T cells.
Furthermore, we demonstrate that helper T cells deficient in the Dnmt3a alternative
transcript, Dnmt3a2, undergo de novo methylation at the IFN-y promoter during THI
polarization,
and that IFN-y expression
is inhibited in these T cells.
Collectively,
this
suggests that Dnmt3a is required for efficient de novo methylation of the IFN-y promoter
during TH2 polarization, and that Dnmt3a2 suppresses IFN-y methylation during THI
polarization.
128
Introduction
It is likely that dynamic changes in the methylation pattern of the IFN-y locus
during helper T cell polarization play a role in transcriptional regulation. While there is a
general correlation between expressed genes and promoter hypomethylation and between
silent genes and promoter hypermethylation, the methylation changes at the IFN-y
promoter do not always follow this pattern. Nafve helper T cells do not express IFN-y
and are hypomethylated at the IFN-y promoter and hypermethylated at the contiguous
transcribed region (Winders et al., 2004). Upon activation, these cells can polarize into
either T 1 cells, which express IFN-y, or TH2 cells, in which IFN-y expression is
repressed (Jankovic et al., 2001; Mosmann et al., 1986; Murphy and Reiner, 2002).
During TH1 polarization, the IFN-y promoter proximal transcribed region undergoes a
slow demethylation,
whereas during TH2 polarization
the promoter
becomes
hypermethylated (Winders et al., 2004; Young et al., 1994). It is unclear how these
methylation changes are regulated and which enzymes are involved.
The dynamic methylation changes that occur at the IFN-y promoter during helper
T cell polarization
have been well characterized,
but it is unknown which
methyltransferase enzymes are responsible for this process. There are three known
mammalian methyltransferase genes, Dnmtl, Dnmt3a, and Dnmt3b (Bestor et al., 1988;
Okano et al., 1999; Okano et al., 1998).
hemimethylated
Dnmtl
preferentially
methylates
DNA (Bestor, 1992), and co-localizes with the replication fork
(Leonhardt et al., 1992).
Targeted inactivation
of Dnmtl results in extensive
demethylation of all examined sequences (Li et al., 1992). It is therefore likely that
Dlnmtl is primarily responsible for the maintenance of established methylation patterns
during cell division. While Dnmtl can methylate unmethylated DNA in vitro (Bestor,
1992), there is little evidence that it acts as a de novo methyltransferase in vivo. Dnmtl
deficiency is embryonic lethal in mice (Li et al., 1992), but mice with a T cell specific
deficiency of Dnmtl show impaired T cell development, impaired activation-induced
proliferation, and a disregulation of cytokine gene expression (Lee et al., 2001). This
further demonstrates the importance of CpG methylation in immune function in general,
and cytokine regulation in particular.
129
The methyltransferases Dnmt3a and Dnmt3b show an enzymatic preference for
unmethylated DNA over hemimethylated DNA (Okano et al., 1998), suggesting that
these enzymes function as de novo methyltransferases. These genes have been shown to
be highly expressed in undifferentiated ES cells (Okano et al., 1998), and are largely
responsible for the establishment of the methylation pattern during early development
(Okano et al., 1999). Dnmt3a has broad somatic expression across many tissues, whereas
Dnmt3b expression shows a more tissue-specific pattern (Chen et al., 2002).
While
Dnmt3b is expressed at low levels ubiquitously (Baylin et al., 2001), there is elevated
expression in the testis, ovary, liver, spleen and thymus (Chen et al., 2002). Where mice
deficient for Dnmt3b do not survive embryogenesis, Dnmt3a deficient mice develop to
term, and appear normal at birth, although they quickly become runted and rarely survive
past 4 weeks of age (Okano et al., 1999).
Both the Dnmt3a and Dnmt3b genes are regulated via alternative splicing. At
least 5 separate splice variants of the Dnmt3a gene have been identified (Weisenberger et
al., 2002) (Figure 1A). While the enzymatic activity of most of these splice forms has
not been characterized, they all contain an intact enzymatic domain, and it is thus likely
that they all have some level of enzymatic activity. The only significant difference that
has been demonstrated between these variants is the expression pattern, with transcripts
containing exon lct being preferentially expressed in somatic cells, and transcripts
containing exon 13 being preferentially expressed in embryonic stem cells (Weisenberger
et al., 2002).
In addition, at least 6 different splice variants of Dnmt3b have been
identified (Chen et al., 2002; Hansen et al., 1999; Okano et al., 1998; Xie et al., 1999)
(Figure lB). Two of them, Dnmt3bl and Dnmt3b2, have been shown to have enzymatic
activity in vitro (Okano et al., 1998), whereas Dnmt3b3, Dnmt3b4, Dnmt3b5, and
Dnmt3b6 are all lacking part or all of the enzymatic domain of the protein, and thus
likely do not have enzymatic activity.
In addition to encoding different splice variants, Dnmt3a also has an alternative
intronic promoter, and transcription from this promoter produces the transcriptional
variant Dnmt3a2 (Chen et al., 2002) (Figure 1A). Dnmt3a2 contains most of Dnmt3al,
including the enzymatic regions, but lacks 219 N-terminal amino acids found in
Dnmt3al. Dnmt3a2 shows similar methylation activity in vitro to Dnmt3al (Chen et al.,
130
Dnnt3a Locus
A1
ATG
ATG
H
1
Fr1111I
l1i
1p
1 2
3 4
5
6
*
1ii
7 8 910 11 12
I lI
11
I
13 14 15 16
17
1819
20 21
22
23
1
Dnmt3al
Splice
Variants
~~~1a~~~~~~~~
1
Dnmt3a2
Dnnt3b Locus
13
ATG
12
3
45
6
7
8 9
8
10111213141516171819
10
19
8
10
8
10
2021
22
22
Dnmt3bl
\JK7\Vl~AJ~y\J~v\AJ/~/\/Wv\v/\J/
\VJK1'AA" //V/\\
Splice
Forms
A2A
8\1t
19
\/AJ/\JJ
19
vDnmt3b4
\Vh7//Vv,\1v\vA~V'A~f
8
10
8
10
8
10
V\f,7\YN1
fAVAVAP.V\J\
19
22
22
Dnmt3b2
Dnmt3b3
22
\J7YAJ/v/J/
'\I\/KJ Dnmt3b5
19
22
19
22
AAVI"7~
Dnmt3b6
Figure 1. Alternative Splice forms of Dnmt3a and Dnmt3b
(A) The structure of the mouse Dnmt3a gene and mRNA splice forms, as well as the
Dnmt3a2 alternative transcript. Exons are shown as black bars. An asterisk indicates the
Dnmt3a2-specific exon. (B) The structure of the mouse Dnmt3b gene and mRNA splice
forms. Exons are shown as black bars.
131
2002), but where mice deficient in Dnmt3al alone have a similar phenotype to mice
completely deficient for Dnmt3a, Dnmt3a2 deficient mice appear to be normal (En Li,
personal communication). Where both Dnmt3al and Dnmt3b localize to heterochromatic
regions of the nucleus (Bachman et al., 2001), Dnmt3a2 localizes to euchromatic regions
of the nucleus (Chen et al., 2002). Furthermore, where Dnmt3al is expressed at similar
levels across a broad array of somatic tissues, Dnmt3a2 expression appears to be largely
restricted to the spleen, thymus and testis (Chen et al., 2002).
It is thus clear that
Dnmt3al and Dnmt3a2 play different physiological roles, despite their structural
similarity.
This chapter describes preliminary studies examining the roles of Dnmt3a,
Dnmt3a2 and Dnmt3b in IFN-y promoter methylation changes during helper T cell
polarization. We demonstrate that Dnmt3b is not required for the de novo methylation of
the IFN-y promoter during TH2 polarization.
We also demonstrate that de novo
methylation of the IFN-y promoter is reduced when Dnmt3a is partially deleted from T
cells.
Finally, we demonstrate that T cells deficient in Dnmt3a2 undergo de novo
methylation when activated under both THi and TH2 polarizing conditions. Significantly,
Dnmt3a2 deficient T cells also show reduced IFN-y expression when polarized under TH1
conditions. These results indicate that Dnmt3a plays a role in the de novo methylation of
the IFN-y promoter during TH2 polarization, and that Dnmt3a2 inhibits Dnmt3a-mediated
methylation of the IFN-y promoter during TH1 polarization.
132
Results
° x Mice Have Fewer T cells and a Higher Percentage of Activated
lckCreDnmt3bz`
CD8 + T cells Than Littermates
We acquired mice with a loxP flanked Dnmt3b enzymatic domain (Figure 2) that
expressed Cre recombinase under the control of the ck promoter (ckCreDnmt3b 2 °1 x)
from Dr. En Li of Massachusetts General Hospital.
In these mice, Cre mediated
recombination rendered cells Dnmt3b deficient during the double negative stage of
thymic development. The average cell number in the spleen of the mutant mice was
about 40% lower than in the spleens of littermate Dnmt3b2 °1x mice that did not express
Cre (these mice will be referred to as littermates, p < 0.02) (Figure 3). While the
thymuses and lymph nodes also had consistently fewer cells in the mutant mice than in
the littermates, the difference was not statistically significant (Figure 3). FACS analysis
of the thymuses of mutant mice appeared normal, with similar levels of CD4, CD8
double negative, CD4, CD8 double positive, CD4 single positive, and CD8 single
positive cells when compared to littermate controls (Figure 4).
The spleens of the ckCreDnmt3b2 °1x mice contained a significantly lower
percentage of T cells than their littermates (20.8% versus 31.5% T cells, p < 0.02) (Figure
5A). This difference was also present in the lymph nodes of these mice, with the
lckCreDnmt3b2 °1x lymph nodes consisting of 47.3% T cells, but the littermate lymph
nodes consisting of 64.8% T cells. While there are fewer total cells for both CD4+ T cells
and CD8+ T cells in spleens and lymph nodes of the ckCreDnmt3b2°1x mice, the reduction
is greater in the CD8+ population, which results in a skewing of the ratio of CD4+ T cells
to CD8 + T cells (Figure 5B).
The activation state of the CD4+ T cells as measured by CD45RB and CD62L
expression
was similar in ckCreDnmt3b 2 °1x mice and littermates (Figure 6). However,
gh
there was a significantly higher percentage of CD44hi
CD8+ T cells in the spleens and
lymph nodes of lckCreDnmt3b2 °1x mice (68.2% and 48.6% respectively,
p < 0.03) than in
littermates (41.5% and 29% respectively, p < 0.03) (Figure 7). This indicates that
lckCreDnmt3b2 'ox mice have a greater percentage of activated and memory CD8+ T cells
than do littermates.
133
Dnnt3b 21ox
A
loxP
FG
12
I I 111 I
3
45 6
7
I11 I
8 9
IoxP
I 11111*1
I I
1011121314 15 16171819
2021
22
Dnnt3b 1lox
A I
~~~~~~~~~I
I
|1
12
I
I
3
11
II I
45 6
I
I~
7
oxP
loxP
I 1
Iif Io
IIr
8 9 1011121314 15
I
I
I I
1
II
20 21
22
Figure 2. Cre mediated deletion of Dnmt3b
The structure of the Dnmt3b gene in Dnmt3b2 "° x mice before (2lox) and after (1 lox) Cremediated recombination.
134
250
ill
200
0
T"""
2:S 150
~
Q)
..c
E
::J
z 100
Q)
0
50
0-
Figure 3. There are fewer cells in the lymphoid organs of I ckCreDnmt3b21ox
mice
than Iittermates.
The total cell number from the spleens, lymph nodes, and thymuses of l8-week-old
lckCreDnmt3b21ox mice and littermate controls that do not express Cre (WT). Values are
the averages of three mice. Error bars represent the standard deviation. P values were
determined by the student t-test.
135
Wild Type
Ak
13.6
IckCreDnmt3b
78.
2
Iox
85.2
7.56
,.~~~~~~~~::"'
·
...
3;94
3.66
..·
.
84
r5A5
'......
2.49
J, '
''1
' .
'
,'""1
'rig
81.8
8.52
1.Z:...
0
0'l
; .
.
:
:.
4.73
"'..'"I
7.55
.
... ..
''~'.E
3.01
........
·
;! .... ......
6 78
r'....
.',
,,;,,
' · ·
",'1
·
-....
i
294
,....
CD8
Figure 4. Thymic cell populations are similar between lckCreDnmt3b2 ° x mice and
littermates.
FACS plots indicating the CD4 and CD8 expression status of thymocytes taken from 182 ox mice and littermates that do not express Cre (Wild Type).
week-old ckCreDnmt3b
The two plots from each genotype are representative of 5 mice from each genotype.
136
70
A
60
50
C/)
40
a.>
()
~ 30
?ft.
20
10
o
..•
B
38.6.
.... '
"
.
0.5
.
0.5
38.8 ..
o.
....~\.;~,;,'~'d~;.~.
,':
&.~,,~)~~:::.:
,
.
39.4
.'.
....
J~::.":.t., ..
:~.,+";..
WT
0.95.
59.9
1.27
.58.7
1.63.
24.4
0.25
27.4
0.77
27.9
1.0
IckCreDnmt3b21ox
co
o
o t.85
0.77
CD4
Fi gure 5. There are fewer T cells and a
IckCreDnmt3b21oll mice compared to littermates.
higher CD4 to CDS
ration
in
(A) The percent TCR~ positive cells in spleens and lymph nodes of IckCreDnmt3b21ox
mice and littermates that do not express Cre (WT) as determined by FACS. Error bars
represent standard deviations of three mice analyzed. P values were determined by the
student t-test. (B) FACS plots indicating the CD4 and CD8 expression status of TCR~
positive cells taken from the lymph nodes of 18-week-old IckCreDnmt3b210ll mice and
littermates that do not express Cre (WT). Three mice were analyzed.
137
CD4+ T cells
90
A
80
70
-5, 60
£:
co 50
0:
to
""""
0
0
-;:R.
0
40
30
20
10
0
~
~
S.:Y
~
~flj
~flj
VjQ..
0Q..
~
s:=
~~
s:=
Q(;:'
Q(;:'
B
90
80
70 -
-5,
£:
.....J
60
50
C\J
(0
o
o
eft. 30
20
10
o
Figure 6. The activation status of CD4+ T cells is similar in IckCreDnmt3b2lox
mice
and Iittermates.
(A) The percentage of CD4+ cells that are CD45RB high in spleens and lymph nodes of
IckCreDnmt3b21ox m ice and littermates that do not express Cre (WT) as determined by
FACS. Error bars represent standard deviations of three mice analyzed. (B) The
percentage of CD4+ cells that are CD62L high in spleens and I ymph nodes of
IckCreDnmt3b21ox m ice and lit termates that do not express Cre (WT) as determined by
FACS. Error bars represent standard deviations of three mice analyzed.
138
80
CD8+ T cells
60
g, 50
..c
~ 40
o
() 30
#-
20
10
o
Figure 7. A higher percentage of CD8+ T cells are CD44 high in IckCreDnmt3b21ox
mice than in littermates.
The percentage of CD8+ cells that are C044 high in spleens and lymph nodes of
lckCreDnmt3b21ox m ice and littermates that do not express Cre (WT) as determined by
FACS. Error bars represent standard deviations of three mice analyzed. P values were
determined by the student t-test.
139
Helper T cell Polarization is Normal in Dnmt3b Deficient T cells
We next investigated the role of Dnmt3b in helper T cell polarization.
Naive
CD4+ T cells were purified from the spleens and lymph nodes of either lckCreDnmt3b2 "° x
mice or littermates by FACS. This was followed by in vitro activation using plate-bound
anti-CD3 and anti-CD28 under polarizing conditions. THi polarization was done by
adding recombinant IL-12 and anti-IL-4 antibody to the culture, while TH2 polarization
was done by adding recombinant IL-4 and anti-IFN-y antibody to the culture.
After 6 days under polarizing conditions, cells were stimulated with PMA and
ionomycin and cytokine
secretion
was assayed by FACS.
Cells from both
lckCreDnmt3b2°1 x mice and littermates that were polarized under TH1 conditions
expressed similar levels of IFN-y and did not express IL-4 following stimulation (Figure
8). Cells from both mice polarized under TH2 conditions expressed similar levels of IL-4,
and no IFN-y following stimulation (Figure 8). Thus, Dnmt3b deficiency had no effect
on effector cytokine secretion during helper T cell polarization.
In order to determine the IFN-y promoter methylation status during helper T cell
polarization for the lckCreDnmt3b2° x mice and littermates, genomic DNA was isolated
from naive, TH1 and TH2 cells, bisulfite converted, PCR amplified, cloned and sequenced.
There was no significant difference in the IFN-y locus methylation state between the
mutant mice and the wild-type mice in any of these populations (Figure 9). In naive cells
from both types of mice the IFN-y promoter was hypomethylated while the transcribed
region was hypermethylated. During TH1 polarization, cells from both the mutant mice
and littermates maintained a hypomethylated promoter and had less methylation at both
CpGs of the coding region. During TH2 polarization, cells from both mouse genotypes
underwent a de novo methylation of the promoter, with all CpGs increasing their
methylation
level to varying degrees,
while the transcribed
region remained
hypermethylated. Thus, it is clear that Dnmt3b is not required for de novo methylation of
the IFN-y promoter during TH2 polarization.
140
WT
IckCreDnmt3b210X
Figure 8
•
No Restimulation
B
P/I Restimulation
IFN-g
WT
IckCreDnmt3b210X
•
No Restimulation
B
P/I Restimulation
IFN-g
IL-4
141
Figure 8. Cytokine expression following helper T ce 11polarization is similar in
2 ° x mice and littermates.
lckCreDnmt3b
IFN-y and IL-4 expression by TH1 and TH2 polarized helper T cells from
lckCreDnmt3b2° x mice and littermates that did not express Cre (WT). TH1 cells were
generated by activating FACS purified naive CD4+ T cells with plate-bound anti-CD3
and anti-CD28 antibodies for 6 days in the presence of recombinant recombinant IL-12
and neutralizing anti-IL-4 antibody. TH2 cells were generated by activating FACS
purified naive Balb/c CD4+ T cells with plate-bound anti-CD3 and anti-CD28 antibodies
for 6 days in the presence of recombinant IL-4 and neutralizing anti-IL-12 antibody.
Polarized TH and T. 2 cells were either stimulated with PMA and ionomycin or left
unstimulated for hour and cytokine expression was measured by cytokine recapture
FACS. The open histogram represents stimulated cells while the shaded histogram
represents unstimulated cells.
142
Wild Type
100
80
D
c
0
••
60
+::;
rn
>£.
+-'
Q)
40
~
~0
20
Naive
TH 1
TH2
0-205
-191
-53
-171
-47
-34
+16
+96
CpG
IckCreDnmt3b21OX
100
80
D
c
0
.+:;
m
••
60
'5.
£.
Q)
~
~0
40
20
-205
-191
-171
-53
-47
-34
+ 16
Naive
TH 1
TH2
+96
CpG
21ox
Figure 9. IckCreDnmt3b
m ice and littermates
methylation states during helper T cell polarization.
have similar IFN-y promoter
The percent methylation at individual CpGs of the IFN-)' proximal promoter for na"ive
helper T cells, THI cells and T H2cells generated from either lckCreDnmt3b21ox mice or
littermates that did not express Cre (Wild Type). Naive CD4+ T cells were purified from
the spleens and lymph nodes of the indicated mice by FACS. T HI cells were generated
by activating FACS purified naive CD4+ T cells with plate-bound anti-CD3 and antiCD28 antibodies for 6 days in the presence of recombinant IL-12 and neutralizing antiIL-4 antibody. T H2cells were generated by activating FACS purified naive Balb/c CD4+
T cells with plate-bound anti-CD3 and anti-CD28 antibodies for 6 days in the presence of
recombinant IL-4 and neutralizing anti-IL-12 antibody. The methylation status of the
cells was determined by bisulfite sequencing. Between 18 and 23 clones were analyzed
for each sample. Sequenced clones are from two different PCR amplifications with two
independently isolated DNA samples.
143
The IFN-y Promoter Methylation is Reduced during TH2 Polarization When
Dnmt3a is Partially Deleted
In order to examine the role of Dnmt3a in the methylation changes at the IFN-y
promoter during TH2 polarization we used ER-CreDnmt3a2°xi
2 °x
mice. These mice
contain a transgenic estrogen-receptor/Cre fusion protein as well as Dnmt3a alleles in
which loxP sites flank the enzymatic domain (Figure
OA). When these mice are
exposed to tamoxifen Cre localizes to the nucleus and is thus able to disrupt the floxed
Dnmt3a genes. Unfortunately, treating these mice with tamoxifen in vivo results in
inefficient disruption of the Dnmt3a gene in the peripheral lymphoid organs (data not
shown). Because of this, we used in vitro tamoxifen treatment of isolated cells to disrupt
the Dnmt3a gene.
Naive CD4+ T cells were purified by FACS from the spleens and lymph nodes of
° x/
2 1° x mice and mice that contained ER-Cre transgenes, but were wild
ER-CreDnmt3a2'
type at the Dnmt3a locus (referred to as wild-type). These cells were activated under TH2
polarizating conditions as described above, with the exception that tamoxifen was added
°xI /2 °x cells. After 6 days in
to the littermate cells and a subset of the ER-CreDnmt3a2b
°x
culture, the ER-CreDnmt3a2'
/2 °x
cells that had been cultured with or without tamoxifen
were restimulated with anti-CD3 and anti-CD28 under TH2 polarizing conditions for 6
additional days. The cells' prior tamoxifen treatment was continued during this time.
DNA was isolated from naive, day 6, and day 12 cells and analyzed for disruption of the
Dnmt3a gene by a PCR reaction specific for the loxP flanked Dnmt3a gene (Dnmt3a
21ox), which does not recognize the wild-type locus or the locus following Cre mediated
° cells that were
recombination (Dnmt3a ilox). When DNA from ER-CreDnmt3a2°ox /2 ox
not exposed to tamoxifen was used as a template for this reaction, bands of the same size
and similar intensity were formed whether the DNA was isolated on day 0, 6 or 12
(Figure 11). This suggests that Cre-mediated recombination did not take place in the
absence of tamoxifen. When DNA from ER-CreDnmt3a2° x/2I2°x cells that were exposed to
tamoxifen was used as a template for this reaction, a band of the same size but reduced
intensity relative to the untreated samples was generated by the day 6 sample, whereas
the day 12 sample did not produce a band (Figure 11). This indicates that tamoxifen-
144
A
Dnnt3a 21ox
ATG
ATG
[ lI
II
I1
l2
3II
1l l 2 3 4 5
1
1.
I I
Il
6
ATG
I
Ab
¢"
*
1,
I
I
1-h.. o
I
I
I
I
I
la 2 3 4 5 6
*
I I II
1 III
I
I
I
ATG
I
I
I
I
l
7 8 9 10 1112
IE II I I I I I
I
loxP
I
13 14 15 16
Dnnt3a 1lox
I iu II I _
17
1819
2021
i
I
I
I
I
I
I
I
I
I
I
iL I
I
I
I
I
I
I
I
I
I
I
I
I
I
11 I
I
I
B
23
loxP
I
7 8 910 11 12
22
13 14 15 16
17 19 20 21
I
I
22
23
Dnnt3a2 knock out
ATG
ATG
I
Ft1 I I
I lI
113 la 2
I
I I I
I I I I
3 4 5 6
I 11
I I I
II II II '
7 8 9 10 11 12
I I I
I I I
13 14 15 16
I
17
11
I I I I I I
18 19 20 21 22
23
Figure 10. The Dnmt3a Locus in Dnmt3a mutant mice
(A) The structure of the Dnmt3a gene in Dnmt3a2 "° x mice before (21ox) and after (ox)
Cre-mediated recombination. (B) The structure of the Dnmt3a gene in Dnmt3a2-' - mice.
145
2 lox
IFN-y intron 1
Figure 11. ere mediated recombination
at the dnmt3a locus was still incomplete
after 6 days in culture with tamoxifen, but was co mplete after 12 days in culture
with tamoxifen.
PCR specific for dnmt3a2lo\ DNA was performed on cells obtained from mice with a
wild-type dnmt3a locus (WT), or on cells from ER-CreDnmt3a2Im/2Im mice. Cells were
either FACS purified na'fve CD4+ T cells or TII2 cells that had been polarized for 6 (d6) or
12 (dI2) days in the presence or absence of tamoxifen (Tx). As a control, a PCR specific
for IFN-y intron I was also done on the same samples.
146
induced Cre-mediated recombination was not complete after 6 days in culture, but was
complete after 12 days in culture. As expected, the PCR did not generate bands when
wild-type DNA or no DNA was added to the PCR reaction. As a positive control, a PCR
reaction specific for a portion of IFN-y intron I was done using all DNA samples as
templates. Similar size and weight bands were generated from all the samples, but not
when DNA was withheld from the reaction. The above data indicate that the tamoxifeninduced disruption of Dnmt3a was not complete after 6 days in culture, but was complete
after 12 days in culture.
We next examined the effect of this gradual disruption of the Dnmt3a locus on the
de novo methylation of the IFN-y promoter that occurs during TH2 polarization. DNA
was isolated from naive CD4+ T cells purified from the spleens and lymph nodes of ERCreDnmt3a2"x2 °x mice, and cells from the same mice that had been polarized under TH2
conditions for either 6 or 12 days in the presence or absence of tamoxifen. To control for
non-specific tamoxifen effects, wild-type naive helper T cells were also polarized for 6
days in the presence of tamoxifen. Following TH2 polarization, tamoxifen treated wild° 'X/
2 ° x cells expressed
type cells, and both tamoxifen treated and untreated ER-CreDnmt3a2
IL-4, but not IFN-y upon stimulation (Figure 12), indicating that the TH2 polarization was
successful.
The IFN-y promoter was hypomethylated
and the coding region
'°
hypermethylated for the naive CD4+ T cells taken from both the ER-CreDnmt3a2°ox /2ox
mice and the wild-type mice (Figure 13). In cells that had been polarized for 6 days
under TH2 conditions, de novo methylation had taken place at the IFN-y promoter in cells
°ox /2ox
'° cells, regardless of
taken from the wild-type mice, as well as in ER-CreDnmt3a2'
tamoxifen treatment (Figure 13). On the other hand, in cells that had been TH2 polarized
for 12 days, there was significantly
more methylation in the promoter of ER-
°' x/2 ° x cells that had not been exposed to tamoxifen,
CreDnmt3a2I
than in ER-
° cells that had been exposed to tamoxifen (p < 0.001) (Figure 13). This
CreDnmt3a2° ' 2 'x
suggests that Dnmt3a plays a role in the de novo methylation of the IFN-y promoter
during TH2polarization.
The IFN-y promoter of Dnmt3a2 Deficient Helper T cells Undergoes De Novo
Methylation During Both TH and TH2Polarization
147
WT + Tx
ER-LckDnmt3a2Iox/2Iox
ER-LckDnmt3a2loxl2lox
+ Tx
IFN-g
IL-4
•
No Restimulation
Figure 12. Tamoxifen treatment
polarized cells.
B
P/I Restimulation
has little effect on cytokine expression in T H2
IL-4 expression by T H2polarized helper T cells from ER-CreDnmt3a2IOll/2Ioll
mice and
wild-type mice that were either cultured in the presence or the absence of tamoxifen (Tx).
T H2cells were generated by activating FACS purified na"ive Balb/c CD4+ T cells with
plate-bound anti-CD3 and anti-CD28 antibodies for 6 days in the presence of
recombinant IL-4 and neutralizing anti-IL-12 antibody. Polarized T H2cells were either
stimulated with PMA and ionomycin or left unstimulated for 1 hour and cytokine
expression was measured by cytokine recapture FACS. The open histogram represents
stimulated cells while the shaded histogram represents unstimulated cells.
148
Figure 13
c
0
:+::;
co
...
>.
..c
Day 0
Q)
~
...
c
Q)
u
~
Q)
0....
100
90
80
70
60
50
40
30
20
10
0
-
;::-
,-
-
-
-
fff-
-
f-
II
II
-205
II
-190
-170
II
II
-53
f-
II
-45
-34
16
96
CpG
c
0
".+:J
co
...
~
...
>.
Day 6
..c
Q)
c
Q)
u
~
Q)
0....
100
90
80
70 60
50 40 30
20
10
0
- -
-205
-190
-170
-53
-
- -
-45
-34
16
96
-45
-34
16
96
CpG
c
0
:+::;
co
...
~
...
>.
..c
Day 12
Q)
c
Q)
u
~
Q)
0....
100
90
80
70
60
50
40
30
20
10
0
-
-
-205
D
WT + Tx
[J
-190
-170
-53
CpG
ER-CreDnmt3a2loxl2lox
•
ER-CreDnmt3a2loxl2lox +Tx
149
Figure 13. Less de n ovo methylation occurs in 12 day TH2 polarized ER° 1 2'° x cells that had been cultured in tamoxifen than in cells that had not
CreDnmt3a2x/
been cultured in tamoxifen.
The percent methylation at individual CpGs of the IFN-y proximal promoter for naive
helper T cells, and TH2 cells that had been polarized for 6 or 12 days and generated from
either ER-CreDnmt3a2 1x/2 °1 x mice or wild-type mice. Naive CD4+ T cells were purified
from the spleens and lymph nodes of the indicated mice by FACS. Day 6 TH2 cells were
generated by activating FACS purified naive Balb/c CD4+ T cells with plate-bound antiCD3 and anti-CD28 antibodies for 6 days i n the presence of recombinant IL-4 and
neutralizing anti-IL-12 antibody. For day 12 TH2 cells, the same activation and
polarization steps were performed on day 6 TH2 cells. The methylation status of the cells
was determined by bisulfite sequencing. Eight to ten clones were analyzed for each
sample. Data from two independent experiments is shown. P values were calculated by
the chi-squared test.
150
We next investigated the effect of disruption of Dnmt3a2 on the methylation of
the IFN-y promoter during helper T cell polarization.
The mice used for these
experiments lacked the Dnmt3a2 intronic promoter and the exon that is transcribed in
Dnmt3a2, but not Dnmt3al (Figure 10A). These mice were deficient in Dnmt3a2, but
had functional Dnmt3al.
These mice underwent normal development, had a normal
immune system, and showed no overt phenotype (data not shown).
As before, naive helper T cells were purified from the spleens and lymph nodes of
Dnmt3a2'- mice and wild-type mice with a similar genetic background, and polarized
under THI and TH2 conditions. After 6 days in culture, cells were harvested and assayed
for cytokine expression and IFN-y promoter methylation.
TH2 polarized cells from
Dnmt3a2 deficient mice and from wild-type mice expressed similar levels of IL-4 and
had no IFN-y expression upon stimulation (Figure 14). On the other hand, while almost
all of the TH1 polarized wild-type cells expressed IFN-y upon stimulation, a significant
number of the Dnmt3a2 deficient THI polarized cells that failed to upregulate IFN-y
expression upon stimulation (Figure 14). Neither the THI polarized wild-type cells nor
the TH1 polarized mutant cells expressed IL-4 upon stimulation.
The methylation status of the IFN-y promoter was determined for the above cell
subsets by bisulfite sequencing.
The IFN-y promoter was hypomethylated and the
transcribed region hypermethylated in both the wild-type and the mutant naive helper T
cells (Figure 12). Following TH2 polarization, both the wild-type and the Dnmt3a2
deficient cells had significantly higher methylation at the IFN-y promoter compared to
naive cells (p < 0.001) (Figure 15). Surprisingly, while the IFN-y proximal promoter
remained hypomethylated in wild-type TH
1 cells, it was significantly more methylated in
T 1j polarized I)nmt3a2 deficient cells (p < 0.001) (Figure 15). In fact, the level of
methylation in the TH1 polarized Dnmt3a2 -' - promoter was similar to the level of
methylation in the promoter of TH2 polarized cells. Furthermore, while demethylation of
the transcribed
region occurred during TH1 polarization
in wild-type cells, no
clemethylation occurred in the TH1polarized mutant cells. As methylation of the -53
position CpG has been shown to inhibit promoter function, these methylation data are
consistent with the inhibition of IFN-y expression in Dnmt3a2 deficient THI cells, and
suggest that Dnmt3a2 inhibits IFN-y promoter methylation during TH1 polarization.
151
Figure 14
WT
•
No Restimulation
B
P/I Restimulation
•
No Restimulation
B
P/I Restimulation
IL-4
WT
Dnmt3a2-/-
IFN- y
152
Figure 14. IFN-y expression is suppressed in dnmt3a2 deficient TH cells.
IFN-y and IL-4 expression by THI and TH2 polarized helper T cells from dnmt3a2 -' - mice
and wild-type mice with a similar genetic background. TI cells were generated by
activating FACS purified naive CD4+ T cells with plate-bound anti-CD3 and anti-CD28
antibodies for 6 days in the presence of recombinant IL-12 and neu tralizing anti-IL-4
antibody. TfJ2 cells were generated by activating FACS purified naive Balb/c CD4+ T
cells with plate-bound anti-CD3 and anti-CD28 antibodies for 6 days in the presence of
recombinant IL-4 and neutralizing anti-IL-12 antibody. Polarized TH and TH2 cells were
either stimulated with PMA and ionomycin or left unstimulated for 1 hour and cytokine
expression was measured by cytokine recapture FACS. The open histogram represents
stimulated cells while the shaded histogram represents unstimulated cells.
153
Wild Type
100
90
80
c
70
~
co
60 -
>.
.c.
+-'
50
0
Q)
~
40 -
-;1?.
0
30
-
D
Naive
•
TH1
•
TH2
D
Naive
20
10
0
-205
-190
-170
-53
-45
-34
+16
+96
CpG
Dnmt3a2-1100
90
80 c:
70 -
-
0
~
co
60 -
>.
.c.
+-'
50
Q)
~
40 -
-;1?.
0
30 -
••
TH 1
-
20 -10 -
TH2
0-205
-190
-170
-53
-45
-34
+ 16
+96
CpG
1
Figure 15. Dnmt3a2- - cells undergo de novo methylation during Till polarization.
The percent methylation at individual CpGs of the IFN-y proximal promoter for na"i"ve
helper T cells, T HI cells and T H2 cells generated from either Dnmt3a2-1- mice or wild-type
mice with a similar genetic background.
Na"ive CD4+ T cells were purified from the
spleens and lymph nodes of t he indicated mice by F ACS. TH I ce lis were generated by
activating FACS purified naYve CD4+ T cells with plate-bound anti-CD3 and anti-CD28
antibodies for 6 days in the presence of recombinant
IL-12 and neutralizing anti-IL-4
antibody" T H2 cells were generated by activating FACS purified na"i"ve Balb/c CD4+ T
cells with plate-bound anti-CD3 and anti-CD28 antibodies for 6 days in the presence of
recombinant IL-4 and neutralizing
anti-IL-12 antibody.
The methylation status of the
cells was determined by bisulfite sequencing.
Between 18 and 28 clones were analyzed
for each sample. Sequenced clones are from two different peR amplifications with two
independently
isolated DNA sa mples.
P values were determined using chi-squared
analysis.
154
Discussion
While the roles of the de novo methyltransferases Dnmt3a and Dnmt3b have been
examined in embryonic stem cells (Li et al., 1992; Okano et al., 1999; Sakai et al., 2004),
their rolls in the methylation changes that occur in somatic cells of mature organisms
remains largely unknown. By studying the dynamic methylation changes that occur at
the IFN-y promoter during in vitro polarization we are able to examine how de novo
methyltransferases are involved in this process.
This was done by inducing the
polarization of naive helper T cells that were deficient for either Dnmt3a or Dnmt3b, and
observing how the deficiency effects methylation changes at the IFN-y locus. The use of
specific conditional
mutant mice was necessitated
by the severe phenotype of
conventional methyltransferase knockouts. Both Dnmt3a and Dnmt3b play important
roles during embryonic development; a complete knock out of Dnmt3b is embryonic
lethal, while mice that lack Dnmt3a usually die within 4 weeks of birth (Li et al., 1992).
In order to examine the role of Dnmt3b in IFN-y de novo methylation, we used
lckCreDnmt3b2 ° x mice, which were Dnmt3b deficient only in T cells.
The slightly
reduced total T cell number, especially among CD8+ T cells in the lckCreDnmt3b2°1x
mice, suggest that Dnmt3b may also play a roll in T cell development. It is possible that
the high percentage of CD44high CD8 + T cells in the lckCreDnmt3b2°ox mice is a result of
homeostatic proliferation of peripheral CD8+ T cells. This would be induced by the
reduced number of peripheral CD8+ T cells, and would result in a population of CD44h i gh
homeostatic memory CD8+ T cells. Regardless, the nearly complete peripheral lymphoid
compartment and overall good health of the lckCreDnmt3b2 °x mice suggest that Dnmt3b
is largely dispensable in peripheral T cells.
Immunodeficiency,
Centromeric
region instability,
and Facial anomalies
syndrome (ICF) is a human disease that is associated with mutations in the Dnmt3b gene
(Hansen et al., 1999). The major symptoms of ICF include immune dificiency, facial
abnormalities, global DNA hypomethylation, and chromosomal abnormalities localized
mostly to centromeric heterochromatin (Ehrlich, 2003). This correlates well with the
observation that Dnmt3b is necessary for the establishment of the hypermethylation of
centromeric satellite repeats during murine embryonic development (Okano et al., 1999).
The immunodeficiency symptom associated with ICF can take several forms. Most, but
155
not all, ICF patients have agammaglobulinemia resulting from reduced levels of two or
three classes of immunoglobulins (Ehrlich, 2003). In addition, some ICF patients have
reduced T cell and/or B cell levels, and defects in lymphocyte activation (De Ravel et al.,
2001). Most people diagnosed with ICF die during infancy or early childhood as a result
of uncontrolled infection (Brown et al., 1995).
ICF is a rare recessive disease, and fewer than 50 ICF cases have been reported
since the disease was identified during the 1970s (Ehrlich, 2003). The most common
Dnmt3b mutations found in ICF patients are misense mutations of the conserved DNMT
motifs that encode the methyltransferase catalytic domain (Gowher and Jeltsch, 2002).
As complete Dnmt3b deficiency in mice is embryonic lethal, it is likely that there is some
residual activity from the Dnmt3b alleles of ICF patients. This is supported by the fact
that no ICF patients have been identified carrying two Dnmt3b null alleles (Ehrlich,
2003). Furthermore, recombinant Dnmt3b protein carrying missense mutations found in
several ICF patients have diminished, but residual, methyltransferase activity (Gowher
and Jeltsch, 2002). It is possible that the variability in symptom severity in ICF patients
is a result of differences in levels of residual Dnmt3b activity.
As the lckCreDnmt3b2 "° x mice examined in this study were overtly normal and
had a relatively intact immune system, it is unlikely that ICF syndrome is a result of
Dnmt3b deficiency in peripheral T cells. This is not surprising considering that the
majority of ICF patients display humoral immunodeficiency, while only a few show
defects in the T cell compartment (Ehrlich, 2003). It is possible that mice deficient with a
13 cell Dnmt3b deficiency would have an immune phenotype similar to that of ICF
patients.
Furthermore, patients with ICF syndrome have reduced Dnmt3b activity
throughout development and in all cells, while the lckCreDnmt3b2 °1x mice are only
I)nmt3b-deficient in thymocytes and peripheral T cells. Thus, Dnmt3b hypomorph mice
would likely be more suitable ICF model organisms than the lckCreDnmt3b2 ° x mice.
In vitro polarization of helper T cells from lckCreDnmt3b2 °1x mice resulted in the
expected cytokine secretion patterns and normal methylation changes at the IFN-y locus
(Table 1). This indicates that Dnmt3b is not required for de novo methylation of the IFNy promoter during TH2 polarization. However, it remains possible that Dnmt3b plays a
role in this methylation event, with other enzymes compensating for its deficiency.
156
Table
1.
Summary of IFN-y promoter and transcribed region methylation status in helper T cell
populations of wild type and methyltransferase deficient cells.
Naive Helper T cells
TH
TH2 Cells
Cells
Genotype
promoter
trans. reg.
promoter
trans. reg.
promoter
trans. reg.
Wild type
Hypo
Hyper
Hypo
Hypo
Hyper
Hyper
Dnmt3b -' -
Hypo
Hyper
Hypo
Hypo
Hyper
Hyper
Dnmt3a -' -
ND
ND
ND
ND
Hypo
Hyper
Dnmt3a2-' -
Hypo
Hyper
Hyper
Hyper
Hyper
Hyper
Naive helper T cells, THI cells and TH2 cells were generated as described in the experimental
procedures. Wild type cells were obtained from Balb/c mice. Dnmt3b -/ - cells were obtained from
lckCreDnmt3b2 °ox mice. Dnmt3a/-' - cells were obtained from ERCreDnmt3a2°joX /2 ° x mice and treated
with tamoxifen for 12 days. Dnmt3a2 -/- cells were obtained from Dnmt3a2 -' - mice.
Abbreviations: Promoter: IFN-y proximal promoter, trans. reg.: IFN-y promoter proximal transcribed
region, Hypo: hypomethylated,
Hyper: hypermethylated,
ND: no data
157
To examine the roll of Dnmt3a in the de novo methylation of the IFN-y locus, we
°x
used ER-CreDnmt3a2I
/2 °1
x mice. These mice express Cre in response to tamoxifen.
Unfortunately, administering tamoxifen to live mice failed to efficiently delete Dnmt3a in
peripheral lymphoid organs. Because of this, we induced the Cre recombination event in
vitro by including tamoxifen in the culture during activation and polarization. Under
these conditions, the deletion of Dnmt3a was slow, with Dnmt3a still detectable by PCR
after 6 days. The fact that the band generated by the day 6 DNA was less intense than the
bands from samples that had not received tamoxifen suggests that some recombination
took place during the first six days. Furthermore, it is possible that some active Dnmt3a
protein persists for a time following the disruption of the Dnmt3a locus due to the
stability of the protein and its RNA message.
It is interesting to note that in the
methylation analysis of the day 6 DNA from the tamoxifen treated cells, there is no
reduction in the methylation percentage of the CpG located at the -53 position compared
to the untreated samples, but there is less methylation at the other CpGs within the
proximal promoter. As the -53 CpG has been shown to become methylated faster than
the other CpGs, it is possible that the -53 CpG became methylated before the Dnmt3a
was disrupted, while the other CpGs did not. Regardless, the methylation data from the
day 6 cells are inconclusive.
We also examined the methylation
state of the tamoxifen treated ER-
CreDnmt3a2° x2 '°x cells after 12 days under TH2 polarizing conditions. There was more
methylation
at the IFN-y promoter in these cells than in naive cells, but there was also
significantly less methylation at the promoter in cells cultured with tamoxifen than
without.
While the low level of IFN-y promoter methylation in these cells may be
residual methylation from before the Dnmt3a was disrupted, it is also possible that
another methyltransferase is also involved in the methylation of the promoter during TH2
polarization. Despite this, the data suggests that Dnmt3a plays a roll in the de novo
methylation event that takes place during TH2 polarization.
Although Dnmt3a mice are viable at birth, they usually die within 4 weeks (Li et
al., 1992). The exact cause of death in these mice is unclear, but it is interesting to note
that in addition to being severely runted Dnmt3a -' - mice also have an abnormal intestinal
phenotype, including intestinal bleeding and an abnormal morphology (En Li, personal
158
communication).
This phenotype a more severe form of the phenotype seen in
inflammatory bowel disease, and specifically Crohn's disease (Bouma and Strober, 2003;
Cobrin and Abreu, 2005). This is interesting because one cause of inflammatory bowl
disease is thought to be an overly robust immune response to the indigenous microbes in
the gut (Cobrin and Abreu, 2005). Furthermore, patients with Crohn's disease express
high levels of TH1 cytokines, and especially IFN-y, in the inflamed regions of the
intestine (Bouma and Strober, 2003). Thus, it is possible that the intestinal phenotype
present in the young Dnmt3a -' - mice may be a result of the disregulation of IFN-y and an
overly robust THI response to intestinal microbes.
The most surprising result obtained during this study was that Dnmt3a2 -' - cells
undergo de novo methylation at the IFN-y locus during both TH1 and TH2 polarization
(Table 1). Furthermore, Dnmt3a2 deficient T cells fail to undergo any demethylation of
the coding region during THI1 differentiation.
This suggests that Dnmt3a2 inhibits
methylation of the IFN-y promoter during THI polarization, and is required for the
demethylation of the transcribed region during this same process. This is particularly
startling because it has been shown that Dnmt3a2 has methylation activity both in vitro
and in vivo (Chen et al., 2002), so the result of Dnmt3a2 deficiency would be expected to
be less methylation, rather than more.
There are a number of mechanisms that could explain this result. First, it is
possible that there are alternative splice variants of Dnmt3a2 that lack the enzymatic
region of the gene. While no alternative splicing of Dnmt3a2 has yet been found, the
dnmt3 family member Dnmt3b has at least 4 splice variants that lack enzymatic activity
(Chen et al., 2002; Hansen et al., 1999; Xie et al., 1999). Furthermore, Dnmt3a2
expression has not been closely examined in somatic cells, so if Dnmt3a2 splice variants
exist in these tissues it is unlikely they would have been discovered. A Dnmt3a2 splice
variant lacking enzymatic activity could inhibit methylation of the IFN-y promoter by
preventing Dnmt3a from binding to factors that are either required for targeting to the
IFN-y locus, or required for robust enzymatic activity. One example of such a factor is
Dnmt3L, which is able to directly bind both Dnmt3a and Dnmt3a2 (Sakai et al., 2004),
and is required in vivo for the full enzymatic activity of these methyltransferases (Hata et
al., 2002). Even if there is no inactive splice variant present, this type of mechanism
159
could still take place if Dnmt3a, but not Dnmt3a2 is able to be targeted to the IFN-y
locus. If they compete for common factors necessary for their methyltransferase activity,
this would also restrict Dnmt3a's ability to methylate the IFN-y promoter.
There are also potential indirect mechanisms that could account for Dnmt3a2's
inhibition of IFN-y promoter methylation. For example, if a factor that is responsible for
the targeting of Dnmt3a to the IFN-y promoter following T cell activation is repressed
during TH polarization by the methylation of it's own promoter by Dnmt3a2, the
resulting phenotype would be consistent with what was observed. Unfortunately, it is
impossible to determine the actual mechanism by which Dnmt3a2 prevents IFN-y
promoter methylation during TH1 polarization without further studies.
In this study we demonstrated that Dnmt3b is not required for the de novo
methylation of the IFN-y promoter during TH2 polarization, that Dnmt3a is probably
involved in this process, and that Dnmt3a2 repressed promoter methylation during TH1
polarization. While further studies are required to confirm these findings and to elucidate
the mechanism of regulation of these methyltransferases, knowing the identity of the
methyltransferases involved in this de novo methylation process is an important and
necessary first step.
160
Materials and Methods
Mouse lines and cell culture
LckCreDnmt3b2 °"x mice, ER-CreDnmt3a2°
' /2°1 x
mice and Dnmt3a2 -' mice were all
the generously provided of Dr. En Li of Massachusetts General Hospital. All mice were
on a mixed 129/C57B16 genetic background. Male and female mice between the ages of
10 and 24 weeks were used for all experiments. For controls, Dnmt3b2°ox mice, ER-Cre
mice and 129/C57B16mixed genetic background mice were used, and were also provided
by Dr. En Li.
Primary nafve CD4 + T cell populations (CD4 +, CD45RBh i gh , CD62Lh
gh
, CD44 ° w)
were purified by flow cytometry from lymph nodes and spleens of the indicated mice. All
antibodies used in flow cytometry were obtained from BD Pharmingen, San Diego, CA.
Purity of all cell populations was greater than 97%.
Sorted naive CD4+ T cells were stimulated at a density of 1 x 106/ml with 5 tg/ml
plate-bound anti-CD3 and 5 tg/ml plate-bound anti-CD28 in the presence of polarizing
cytokines. TH1 polarized cultures contained 5 ng/ml recombinant IL-12 and 10 [tg/ml
neutralizing anti-IL-4 antibody (BD Pharmingen), while TH2 polarized cultures contained
25 ng/ml recombinant
IL-4 and 10 [tg/ml neutralizing anti-IFN-y antibody (BD
Pharmingen). After 24 hours, 40 U/ml IL-2 was added to all cultures. Cultures were
washed and fresh cytokines were added every 3 days and restimulated under the above
conditions every 6 days.
When tamoxifen was added to the culture, it was at a
concentration of 100 nM. The following PCR primers were used to detect Dnmt3a2"°x:
Forward:
CTGTGGCATCTCAGGGTGATGAGCA
and
Reverse:
AAGCCTCAGGCCCTCTAGGCAAGAT.
Cytokine Recapture Staining
Cytokine recapture staining was done using either the Mouse IFN-y Secretion
Assay kit or the Mouse IL-4 Secretion Assay kit (both from Miltenyi Biotec, Auburn,
CA) according to the manufacturer's instructions. Briefly, T cell populations were either
restimulated with 5 ng/ml PMA and 100 ng/ml ionomycin in 10% FCS RPMI media or
161
cultured in 10% FCS RPMI media without PMA or ionomycin for 2 or 3 hours. Cells
were washed and stained in 100 R15% FCS RPMI containing 10 tl of the catch antibody
in a 50 ml conical tube on ice. After 5 minutes, 30 ml of warm 5% FCS RPMI was added
to the tubes, and cells were incubated with constant rotation at 37 degrees for 45 minutes.
Cells were washed with cold 1.5 mM EDTA 0.5% BSA PBS buffer and stained in a 100
gl volume with 10 Itldetection antibody and 1 ll APC conjugated anti-IL-4 antibody.
Cell were cultured on ice for 15 minutes, washed, and analyzed by FACS.
Bisulfate Conversion and Methylation Analysis
Methylation analysis was done by bisulfite conversion of genomic DNA using the
CpGenome Universal DNA Modification Kit (Chemicon International, Temecula, CA)
according to the manufacturer's
amplified
using
the
instructions.
The IFN-y promoter was then PCR
following
primers:
TAGAGAATTTTATAAGAATGGTATAGGTGGGTAT
and
Forward:
Reverse:
CCATAAAAAAAAACTACAAAACCAAAATACAATA. The initial annealing
temperature for the PCR reaction was 56 degrees.
The annealing temperature was
reduced at a rate of 0.5 degrees per cycle until an annealing temperature of 47 degrees
was reached.
The PCR reaction was run at this annealing temperature for 25 more
cycles, followed by a 5-minute elongation step. The PCR product was cloned using the
TOPO TA Cloning kit (Invitrogen, Carlsbad, CA), according to the manufacturer's
instructions, and clones were sequenced using the included primers.
Acknowledgements
I thank Dr. En Li for the mouse strains; Dr. Taiping Chen for advice about the
Dnmt3a2-' - mice; and Wendy Marston for help with mouse work.
162
References
Bachman, K. E., Rountree, M. R., and Baylin, S. B. (2001). Dnmt3a and Dnmt3b are
transcriptional repressors that exhibit unique localization properties to heterochromatin. J
Biol Chem 276, 32282-32287.
Baylin, S. B., Esteller, M., Rountree, M. R., Bachman, K. E., Schuebel, K., and Herman,
J. G. (2001). Aberrant patterns of DNA methylation, chromatin formation and gene
expression in cancer. Hum Mol Genet 10, 687-692.
Bestor, T., Laudano, A., Mattaliano, R., and Ingram, V. (1988). Cloning and sequencing
of a cDNA encoding DNA methyltransferase of mouse cells. The carboxyl-terminal
domain of the mammalian enzymes is related to bacterial restriction methyltransferases. J
Mol Biol 203,971-983.
Bestor, T. H. (1992). Activation of mammalian DNA methyltransferase by cleavage of a
Zn binding regulatory domain. Embo J 11, 2611-2617.
Bouma, G., and Strober, W. (2003). The immunological and genetic basis of
inflammatory bowel disease. Nat Rev Immunol 3, 521-533.
Brown, D. C., Grace, E., Sumner, A. T., Edmunds, A. T., and Ellis, P. M. (1995). ICF
syndrome (immunodeficiency, centromeric instability and facial anomalies):
investigation of heterochromatin abnormalities and review of clinical outcome. Hum
Genet 96, 411-416.
Chen, T., Ueda, Y., Xie, S., and Li, E. (2002). A novel Dnmt3a isoform produced from
an alternative promoter localizes to euchromatin and its expression correlates with active
de novo methylation. J Biol Chem 277, 38746-38754.
Cobrin, G. M., and Abreu, M. T. (2005). Defects in mucosal immunity leading to Crohn's
disease. Immunol Rev 206, 277-295.
De Ravel, T. J., Deckers, E., Alliet, P. L., Petit, P., and Fryns, J. P. (2001). The ICF
syndrome: new case and update. Genet Couns 12, 379-385.
Ehrlich, M. (2003). The ICF syndrome, a DNA methyltransferase 3B deficiency and
immunodeficiency disease. Clin Immunol 109, 17-28.
Gowher, H., and Jeltsch, A. (2002). Molecular enzymology of the catalytic domains of
the Dnmt3a and Dnmt3b DNA methyltransferases. J Biol Chem 277, 20409-20414.
Hansen, R. S., Wijmenga, C., Luo, P., Stanek, A. M., Canfield, T. K., Weemaes, C. M.,
and Gartler, S. M. (1999). The DNMT3B DNA methyltransferase gene is mutated in the
ICF immunodeficiency syndrome. Proc Natl Acad Sci U S A 96, 14412-14417.
163
Hata, K., Okano, M., Lei, H., and Li, E. (2002). Dnmt3L cooperates with the Dnmt3
family of de novo DNA methyltransferases to establish maternal imprints in mice.
Development 129, 1983-1993.
Jankovic, D., Liu, Z., and Gause, W. C. (2001). Thl- and Th2-cell commitment during
infectious disease: asymmetry in divergent pathways. Trends Immunol 22, 450-457.
Lee, P. P., Fitzpatrick, D. R., Beard, C., Jessup, H. K., Lehar, S., Makar, K. W., PerezMelgosa, M., Sweetser, M. T., Schlissel, M. S., Nguyen, S., et al. (2001). A critical role
for Dnmtl and DNA methylation in T cell development, function, and survival.
Immunity 15, 763-774.
Leonhardt, H., Page, A. W., Weier, H. U., and Bestor, T. H. (1992). A targeting sequence
directs DNA methyltransferase to sites of DNA replication in mammalian nuclei. Cell 71,
865-873.
Li, E., Bestor, T. H., and Jaenisch, R. (1992). Targeted mutation of the DNA
methyltransferase gene results in embryonic lethality. Cell 69, 915-926.
Mosmann, T. R., Cherwinski, H., Bond, M. W., Giedlin, M. A., and Coffman, R. L.
(1986). Two types of murine helper T cell clone. I. Definition according to profiles of
lymphokine activities and secreted proteins. J Immunol 136, 2348-2357.
Murphy, K. M., and Reiner, S. L. (2002). The lineage decisions of helper T cells. Nat
Rev Immunol 2, 933-944.
Okano, M., Bell, D. W., Haber, D. A., and Li, E. (1999). DNA methyltransferases
Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development.
Cell 99, 247-257.
Okano, M., Xie, S., and Li, E. (1998). Cloning and characterization of a family of novel
mammalian DNA (cytosine-5) methyltransferases. Nat Genet 19, 219-220.
Sakai, Y., Suetake, I., Shinozaki, F., Yamashina, S., and Tajima, S. (2004). Co-
expression of de novo DNA methyltransferases Dnmt3a2 and Dnmt3L in gonocytes of
mouse embryos. Gene Expr Patterns 5, 231-237.
Weisenberger,
). J., Velicescu, M., Preciado-Lopez,
M. A., Gonzales, F. A., Tsai, Y. C.,
Liang, G., and Jones, P. A. (2002). Identification and characterization of alternatively
spliced variants of DNA methyltransferase 3a in mammalian cells. Gene 298, 91-99.
Winders, B. R., Schwartz, R. H., and Bruniquel, D. (2004). A distinct region of the
murine IFN-gamma promoter is hypomethylated from early T cell development through
mature naive and Thl cell differentiation, but is hypermethylated in Th2 cells. J Immunol
173, 7377-7384.
164
Xie, S., Wang, Z., Okano, M., Nogami, M., Li, Y., He, W. W., Okumura, K., and Li, E.
(1999). Cloning, expression and chromosome locations of the human DNMT3 gene
family. Gene 2.36, 87-95.
Young, H. A., Ghosh, P., Ye, J., Lederer, J., Lichtman, A., Gerard, J. R., Penix, L.,
Wilson, C. B., Melvin, A. J., McGurn, M. E., and et al. (1994). Differentiation of the T
helper phenotypes by analysis of the methylation state of the IFN-gamma gene. J
Immunol 153, 3603-3610.
165
Chapter 4:
Further Discussion and Future Direction
166
In this thesis, we examined the methylation changes that occur at the IFN-y
proximal promoter. We observed both a strong correlation between IFN-y transcription
and the methylation state of the promoter and the contiguous transcribed region, and that
these regions underwent dynamic methylation changes during helper T cell polarization.
Interestingly, we noted that the CpGs within the IFN-y promoter do not undergo these
methylation changes at the same rate, with the CpG located at the -53 position becoming
hypermethylated fastest during TH2 polarization. This observation led us to examine the
importance of this rapid methylation event.
In doing so, we demonstrated that
methylation of the -53 CpG is sufficient to inhibit both site binding by members of the
AP1/CREB/ATF family of transcription factors, and promoter function. By combining
these observations with previous results we formulated a model that explains the
relationship between -53 CpG methylation and API/CREB/ATF factor binding at the
IFN-y proximal promoter during helper T cell polarization.
This was discussed in
Chapter 2.
We also examined the involvement of the known de novo methyltransferases in
the methylation changes that occur during helper T cell differentiation. In these studies
we demonstrated that Dnmt3a deficiency, but not Dnmt3b deficiency, inhibits de novo
methylation of the IFN-y promoter during TH2 polarization. Furthermore, we made the
surprising observation that deficiency of the transcriptional variant Dnmt3a2 results in
aberrant methylation of the IFN-y promoter during TH1Ipolarization and inhibition of
IFN-y expression in TH 1 cells. These results were presented in the appendix.
While the findings presented in this thesis further our understanding of the
regulation of IFN-y transcription by promoter methylation, they also raise many questions
and suggest avenues of further investigation. This chapter will focus on some of these
unresolved issues.
IFN-y Locus Methylation
There is a strong correlation between the methylation state of the IFN-y promoter
and a cell's ability to transcribe IFN-y. Cells that do not transcribe IFN-y and are unable
to differentiate into IFN-y transcribing cells are hypermethylated at both the proximal
promoter and at the contiguous transcribed region. Cells that do not transcribe IFN-y but
167
are capable of differentiating into IFN-y transcribing cells are hypomethylated at the
proximal promoter and hypermethylated in the contiguous transcribed region. Finally,
cells that are undergoing active IFN-y transcription are hypomethylated at both the IFN-y
promoter and the contiguous transcribed region. Thus, naive helper T cells undergo a
demethylation of the promoter proximal transcribed region upon becoming IFN-y
expressing TH cells, while they undergo de novo methylation of the IFN-y proximal
promoter as they differentiate into TH2cells and lose their capacity for IFN-y expression.
It is interesting that methylation of the CpGs of the proximal promoter is
regulated differently than the CpGs of the contiguous transcribed region, despite the fact
that they are located within such close proximity. This raises the question of how far
upstream does the methylation status of the IFN-y promoter resemble that of the proximal
region looked at in this study, and how far downstream does the methylation status
resemble that of the promoter proximal transcribed region. Using methylation sensitive
restriction enzymes and semi-quantitative southern-blotting, a recent study has indicated
that the CpG located at the -297 position of the IFN-y promoter is hypomethylated in
naive T cells, remains hypomethylated
in TH1 cells, but becomes hypermethylated
in TH2
cells (Winders et al., 2004). This correlates well with the proximal promoter CpGs. On
the other hand, using the same technique, the CpG located at the -380 position of the
IFN-y
promoter was hypermethylated
hypermethylated
during both TI
in naive helper T cells, and remained
and TH2 polarization
(Winders et al., 2004).
This
would suggest that there is a change in CpG methylation regulation between the -297
CpG and the -380 CpG. It would be interesting to determine if other areas that are
known to contribute to IFN-y transcription are hypomethylated in naive helper T cells, or
if the proximal promoter is unique in this respect.
Looking in the other direction, it is unclear how far into the transcribed region the
methylation pattern matches that of the promoter proximal region examined in this thesis.
A region of intron
is hypermethylated in naive helper T cells, and undergoes a gradual
clemethylationduring TH polarization with the upstream CpGs becoming demethylated
faster than the downstream (Hutchins et al., 2005). Whether this indicates a continuation
of the methylation changes observed at the promoter proximal transcribed region or a
separate region that has a similar methylation pattern is unknown.
168
Regulation of Methylation Changes at the IFN-y locus
This thesis describes dynamic changes that occur in the IFN-y proximal promoter
region during helper T cell polarization. We demonstrated that de novo methylation
occurs at the proximal promoter when cells are activated under TH2 polarizing conditions
but that there is no methylation change when cells are activated without polarizing
cytokines, in the presence of neutralizing anti-IFN-y and anti-IL-4 antibodies.
This
suggests that the signal for methylation of the IFN-y promoter is downstream of IL-4
receptor engagement.
As many of the downstream signals of the IL-4 receptor are known, it should be
possible to determine which signals are important for the induction of IFN-y methylation.
This could be done by ectopically expressing members of this signaling pathway, such as
Stat6, GATA-3, and c-Maf, in IL-4 deficient naive helper T cells activated under neutral
conditions, and then monitoring changes in the methylation status of the IFN-y promoter.
As all of the above factors induce expression of IL-4 (Ho et al., 1996; Lee et al., 2000;
Ouyang et al., 1998), this experiment has to be done in either IL-4 or IL-4 receptor
deficient cells in order to prevent indirect effects.
Induction of IFN-y promoter
methylation by the ectopically expressed factor would indicate that the methylation event
is downstream of its signal.
A similar and complementary experiment could be done using cells deficient in
factors downstream of the IL-4 receptor, either by conventional gene disruption or
through the use of siRNA technologies.
In this experiment the cell containing these
disruptions would be polarized under TH2 conditions and the methylation status of the
IFN-y promoter would again be determined. In this case, a lack of methylation would
indicate that the disrupted gene is necessary for the de novo methylation event.
Methyltransferase Involvement in the Methylation Changes During Helper T cell
Polarization
Our studies using methyltransferase deficient cells have indicated that Dnmt3a,
but not Dnmt3b is involved in the de novo methylation of the IFN-y promoter during TH2
polarization.
We also demonstrated that Dnmt3a2 deficiency results in the de novo
methylation of the IFN-y promoter during TH1 polarization, indicating that Dnmt3a2
169
normally inhibits IFN-y promoter methylation in these cells. How these factors interact
and are regulated remains unknown.
It is surprising that Dnmt3a2 inhibits the de novo methylation of the IFN-y
promoter during TH2 polarization because Dnmt3a2 has methyltransferase activity (Chen
et al., 2002). As both Dnmt3al and Dnmt3b are regulated by alternative splicing (Chen
et al., 2002; Hansen et al., 1999; Okano et al., 1998; Weisenberger et al., 2002; Xie et al.,
1999), it is possible that there are alternative splice forms of Dnmt3a2 as well. Thus, it
could be that the inhibition of methylation during TH polarization is not mediated by full
Dnmt3a2 protein, but by a Dnmt3a2 splice variant that lacks enzymatic activity. For this
reason, it is important to assay for alternatively spliced forms of Dnmt3a2 in the
polarizing T cell populations. Furthermore, the methylation inhibiting activity of full
Dnmt3a2 could be tested by ectopically expressing Dnmt3a2 in T cells from Dnmt3a2
deficient mice that are undergoing T
polarization.
If this restores methylation
inhibition is in these cells, it would be a strong indication that the full Dnmt3a2 protein is
capable of the methylation inhibition. This assay could also be done with any newly
found splice variants in order to test their importance in IFN-y promoter methylation
inhibition.
It is possible that methyltransferases are regulated at the level of transcription
during helper T cell polarization. One hypothetical mechanism for this would be that
upregulation of Dnmt3al expression occurs following TCR engagement independent of
polarizing conditions, while Dnmt3a2 levels are high during TH polarization, but low
during TH2 polarization. Thus, Dnmt3al could methylate the IFN-y promoter in TH2
cells, but would be inhibited by Dnmt3a2 in TH1 cells. This hypothesis could be tested
by assaying for methyltransferase
RNA in naive helper T cells, TH cells and TH2 cells by
RT-PCR. In order to confirm these results and to test for post-transcriptional mRNA
regulation, it would be important to also test these cell populations for methyltransferase
protein by western blot.
Considering that CpG methylation is a general mechanism for gene regulation
(Jeltsch, 2002), and that there are many genes that are either activated or silenced during
TH1 and TH2 polarization (Murphy and Reiner, 2002), it is unlikely that altering
methyltransferase levels is the primary means of controlling methylation of the IFN-y
170
promoter. IFN-y promoter methylation is far more likely to be regulated by factors that
are involved in targeting methyltransferases to the locus. The presence of the different
methyltransferases at the IFN-y promoter during helper T cell polarization could be
assayed through chromatin immunoprecipitation
(ChIP) assays.
Similarly, co-
immunoprecipitation assays could be used to investigate the composition of
methyltransferase-containing complexes in naive and polarizing cells. This assay may
reveal that factors known to bind to the IFN-y promoter are present in the
methyltransferase containing complexes, which would suggest that such factors play a
role in the targeting of the methyltransferase
to the IFN-y promoter.
While the specific pattern of methylation change that occurs at the IFN-y locus
during helper T cell polarization has not yet been reported at other genes, it is likely that
IFN-y is not unique in its use of methylation as a suppression mechanism during helper T
cell polarization. It is possible that other genes that are transcribed in THI cells but
suppressed in TH2 cells, such as T-bet or Hlx, could share regulation mechanisms with
IFN-y. Thus, it would be interesting to examine the role that promoter CpG methylation
may play in these genes. If they do share methylation phenotypes during helper T cell
differentiation, a comparison of promoter elements and known common transcription
factors could suggest promoter elements and factors important for methyltransferase
recruitment.
Finally, it would be interesting to determine whether the promoter
methylation changes that are observed at the IFN-y locus are unique to IFN-y or whether
it is a common regulatory mechanism in helper T cell polarization.
IFN-y Promoter Methylation and Histone Modification
Methyl-CpG binding proteins are able to recruit histone modification enzymes,
especially histone deacetylases and methyltransferases, to hypermethylated loci (Fujita et
al., 2003; Fuks et al., 2003; Jones et al., 1998; Nan et al., 1998). This suggests that DNA
methylation may suppress transcription of a locus by initiating changes in the higher
order chromatin structure of the region, and in doing so "close" the locus and make it less
accessible to transcription factors. This hypothesis is complicated by the observation that
histone modification, especially the methylation of histone H3 at lysine 9, may drive
I)NA methylation (Lehnertz et al., 2003). Thus, it is unclear whether DNA methylation
171
leads to histone modification, or vice versa.
Furthermore, the importance of this
relationship between DNA methylation and histone modification in vivo has not been
demonstrated. The observation that Dnmt3a may be the methyltransferase responsible
for the de novo methylation changes that occur at the IFN-y promoter provides an
opportunity to further investigate these issues.
Following naive helper T cell activation, the histones at the IFN-y promoter
rapidly become acetylated, independent of polarizing conditions (Avni et al., 2002).
Under TH1 conditions, the histones remain hyperacetylated, but under TH2 conditions the
histones become deacetylated (Avni et al., 2002). This histone deacetylation under TH2
conditions correlates with the de novo methylation that takes place at the IFN-y promoter.
Thus, it is possible that the methylation of the IFN-y promoter is leading to the
recruitment of histone deacetylases, and in doing so, driving the deacetylation of the
region. This hypothesis could be tested by investigating the histone acetylation changes
that occur in Dnmt3a deficient T cells. T cells could be purified from mice with a T cell
specific Dnmt3a deficiency (for example lckCreDnmt3a2l° x mice), and activated under
TH2 conditions.
If, in addition to having a block in the methylation of the IFN-y
promoter, there was also a block in the deacetylation of the locus, it would suggest that
methylation of the IFN-y promoter is an important precursor for deacetylation.
As
Dnmt3a2 -' - mice undergo hypermethylation of the IFN-y promoter under TH1 conditions,
looking at the acetylation changes in THI cells from these mice would reveal whether
promoter methylation is sufficient to drive histone deacetylation. These studies could
enhance our understanding of the relationship between DNA methylation and histone
modification.
The IFN-y Promoter Methylation and Transcription
While we were able to demonstrate some of the downstream effects of IFN-y
proximal promoter de novo methylation that occurs during TH2 polarization, the effects, if
any, of the demethylation that takes place at the contiguous transcribed region remain
unclear.
As IFN-y expression
is initiated prior to demethylation
in most TH1 cells, it is
clear that demethylation is not required for expression. At the IL-4 locus, where a slow
demethylation occurs during TH2 polarization, demethylation is not required for IL-4
172
expression, but there is a correlation between demethylation and cells that express a high
level of IL-4 (Lee et al., 2002). It would be interesting to see if a similar correlation is
also present between IFN-y transcribed region demethylation and a high level of IFN-y
expression in TH.1cells. It has also been demonstrated that fully differentiated helper T
cells are able to express IFN-y independently of T-bet (Martins et al., 2005; Mullen et al.,
2002). This raises the question as to whether this T-bet independence is acquired along
with transcribed region demethylation, or whether they are separate events.
While we demonstrated that methylation at the -53 CpG is sufficient to inhibit
IFN-y promoter function in transient transfection assays, and there is a strong correlation
between IFN-y hypermethylation and a lack of IFN-y transcription, the in vivo effect of
endogenous IFN-y promoter methylation remains unknown.
Helper T cells from
Dnmt3a2 -' - mice undergo de novo methylation at the IFN-y promoter when polarized
under both T, 1 and TH2 conditions and IFN-y expression is inhibited in some of the TH1
polarized cells from these mice. It is likely that the Dnmt3a2-' - cells that fail to express
IFN-y under TH conditions are hypermethylated at the IFN-y promoter. This could be
tested by purifying TH polarized cells that express IFN-y and TH polarized cells that do
not express IFN-y, and comparing the methylation status of the IFN-y promoter,
particularly at the -53 position, between these two populations. A positive correlation
between hypomethylation of the locus and IFN-y expression would suggest that IFN-y
promoter methylation is able to inhibit endogenous promoter function.
Following DNA replication, the CpG methylation pattern is maintained by
Dnmtl, which preferentially methylates hemi-methylated DNA (Bestor, 1992) and is
localized to the replication fork (Leonhardt et al., 1992). This provides a mechanism by
which DNA methylation can act as a long-term suppressor of gene transcription. During
helper T cell polarization there is a point at which the differentiation becomes permanent,
and the cell loses its capacity to change polarization to the opposite phenotype (Grogan et
al., 2001). There are many events that are likely to contribute to this process, including
the downregulation of cytokine receptors, the suppression of important transcription
factors, and higher order epigenetic changes at the cytokine loci. It is likely that DNA
methylation plays role in this process as well. The involvement of IFN-y promoter
methylation in the long-term maintenance of polarization state could be addressed using
173
helper T cells deficient in Dnmt3a because methylation of the IFN-y promoter during TH2
polarization in these cells is inhibited. The long-term suppression of IFN-Yin these cells
could be tested by exposing TH2 polarized helper T cells from these mice to TH1
polarizing conditions.
If IFN-y promoter methylation is necessary to inhibit IFN-y
transcription, exposure to TH1 conditions may induce inappropriate cytokine expression.
As the IL-12 receptor is downregulated in TH2 polarized cells (Murphy and Reiner,
2002), it may be necessary to cross mice lacking T cell expression of Dnmt3a with mice
expressing a transgenic IL-12 receptor in order to see an effect. Regardless, it would be
interesting to determine whether IFN-y promoter methylation plays an important role in
the long-term suppression of the locus.
IFN-y Locus Methylation and Immune Function
Our work has suggested that the methyltransferase Dnmt3a plays an important
role in the de novo methylation of the IFN-y promoter during TH2 polarization and
Dnmt3a2 is required to prevent methylation of the IFN-y promoter during TH1
polarization. Dnmt3a deficient mice die very young (Okano et al., 1999) but, prior to
death, have an intact immune system. While our analysis to date has been on cells with
an inducible deletion of Dnmt3a, it is likely that generation of T cell specific Dnmt3a
deficient mice by breeding the Dnmt3a2 °x gene onto a line that carries a ck-Cre
transgene would be possible. As Dnmt3a2 deficient mice are completely viable and have
no obvious immune phenotype, we would then have available both mice that fail to
methylate the IFN-y promoter during TH2 polarization and mice that aberrantly methylate
the IFN-y promoter during TH1 polarization. This would allow us to study the effect of
IFN-y promoter methylation during helper T cell polarization on the immune response in
viVO.
Using Dnmt3a2 deficient mice and lckCreDnmt3a2I° x mice, the effect of IFN-y
promoter methylation on helper T cell polarization in vivo could be examined. When
mice are immunized with keyhole limpet hemocyanin (KLH) in the presence of CFA
they preferentially mount a KLH specific THI response (Grun and Maurer, 1989). On the
other hand, when mice are immunized with KLH in the presence of alum, the mice
preferentially mount a KLH specific TH2 response (Grun and Maurer, 1989). Comparing
174
the polarization of the immune response of Dnmt3a2 deficient mice with
° x mice and wild-type mice would give an indication as to the effect of
lckCreDnmt3a2"
IFN-y methylation on helper T cell polarization and IFN-y expression in vivo. As DNA
methylation may play a roll in the long-term maintenance of the polarization phenotype,
it would be important to examine secondary memory T cell responses in addition to
primary responses.
Mice that are completely deficient in Dnmt3a and mice that are deficient in
Dnmt3al alone, have similar phenotypes (En Li, personal communication). Both mouse
lines appear normal at birth, but rapidly become runted compared to littermates, and
usually die within a few weeks (Okano et al., 1999). Both strains manifest an obvious
intestinal phenotype that includes structural problems and bleeding. While it has yet to
be fully examined, it is likely that one of the causes of death in these mice is malnutrition
due to these intestinal abnormalities.
The involvement of the immune system in this intestinal phenotype has not yet
been examined, but considering that an uncontrolled immune response to intestinal
microbes is thought to be the main cause of inflammatory bowl disease (IBD) (Bouma
and Strober, 2003), it is worth investigating. Furthermore, IBD, and especially Crohn's
disease, are often associated with the expression of high levels of THi cytokines at the
site of inflammation (Cobrin and Abreu, 2005). As Dnmt3a appears to play a role in
IFN-y suppression, this provides a possible link between any observed IBD phenotype in
these mice and the function of Dnmt3a. In addition to examining the T cell phenotype
and cytokine expression profile in the gut of young Dnmt3a ' - mice, it is possible that an
examination of mice with a T cell specific deficiency in Dnmt3a may reveal a direct role
for helper T cell polarization in this phenotype. It may also be worthwhile to cross
lckCreDnmt3a2I° x mice onto mice that model IBD, such as Mdrla -' mice, to see if a lack
of TH2 IFN-y promoter methylation has an effect on disease progression.
Mouse strains vary in their susceptibility to pathogenic infection by specific
microbes depending upon their predisposition towards either a TH1 response or a TH2
response. For example, Balb/c mice are predisposed to mount a TH2 response, and as a
result are resistant to helminth infection, but vulnerable to infection by Leishmania major
(Heinzel et al., 1988; Himmelrich et al., 1998; Sadick et al., 1986). C57B16, mice, on the
175
other hand, are predisposed to mount a THI response, and as a result are vulnerable to
helminth infection, but resistant to Leishmania infection (Heinzel et al., 1988;
Himmelrich
et al., 1998; Sadick et al., 1986).
It would be interesting
to determine
whether an inability to methylate the IFN-y promoter, or aberrant methylation of the IFN-
y promoter during TH1 polarization has any effect on microorganism susceptibility.
In order to test the effect of IFN-y promoter methylation on the immune response
to pathogenic infection, wild-type mice, Dnmt3a2 deficient mice and ckCreDnmt3a2i° x
mice could be infected with a variety of microorganisms, including the intracellular
parasite Leishmania major, extracellular helminths, and viruses like influenza. The type
of response mounted by the various mouse lines, as well as their ability to clear the
pathogen, could then be monitored. This would give an indication as to the importance
of IFN-y promoter methylation on the primary immune response.
In addition to
following the primary immune response, it would be important to look at the effect of
IFN-y promoter methylation on a secondary immune response.
It is likely, but not
certain, that the Dnmt3a deficient mice will be predisposed toward a TH2 response due to
their methylation of the IFN-y promoter during TH polarization. These mice may be
more susceptible to Leishmania or viral infections, but may be more resistant to helminth
° x mice could be predisposed toward a
infections. It is also possible that lckCreDnmt3a2"
THI response due to their inability to methylate the IFN-y promoter during TH2
polarization. These mice may be more resistant to Leishmania or viral infections, but
may be more vulnerable to helminth infections. These experiments could help elucidate
the importance of IFN-y methylation during helper T cell polarization in the immune
response.
176
References
Avni, O., Lee, D., Macian, F., Szabo, S. J., Glimcher, L. H., and Rao, A. (2002). T(H)
cell differentiation is accompanied by dynamic changes in histone acetylation of cytokine
genes. Nat Immunol 3, 643-651.
Bestor, T. H. (1992). Activation of mammalian DNA methyltransferase by cleavage of a
Zn binding regulatory domain. Embo J 11, 2611-2617.
Bouma, G., and Strober, W. (2003). The immunological and genetic basis of
inflammatory bowel disease. Nat Rev Immunol 3, 521-533.
Chen, T., Ueda, Y., Xie, S., and Li, E. (2002). A novel Dnmt3a isoform produced from
an alternative promoter localizes to euchromatin and its expression correlates with active
de novo methylation. J Biol Chem 277, 38746-38754.
Cobrin, G. M., and Abreu, M. T. (2005). Defects in mucosal immunity leading to Crohn's
disease. Immunol Rev 206, 277-295.
Fujita, N., Watanabe, S., Ichimura, T., Tsuruzoe, S., Shinkai, Y., Tachibana, M., Chiba,
T., and Nakao, M. (2003). Methyl-CpG binding domain 1 (MBD1) interacts with the
Suv39hl-HP1 heterochromatic complex for DNA methylation-based transcriptional
repression. J Biol Chem 278, 24132-24138.
Fuks, F., Hurd, P. J., Wolf, D., Nan, X., Bird, A. P., and Kouzarides, T. (2003). The
methyl-CpG-binding protein MeCP2 links DNA methylation to histone methylation. J
Biol Chem 278, 4035-4040.
Grogan, J. L., Mohrs, M., Harmon, B., Lacy, D. A., Sedat, J. W., and Locksley, R. M.
(2001). Early transcription and silencing of cytokine genes underlie polarization of T
helper cell subsets. Immunity 14, 205-215.
Grun, J. L., and Maurer, P. H. (1989). Different T helper cell subsets elicited in mice
utilizing two different adjuvant vehicles: the role of endogenous interleukin 1 in
proliferative responses. Cell Immunol 121, 134-145.
Hansen, R. S., Wijmenga, C., Luo, P., Stanek, A. M., Canfield, T. K., Weemaes, C. M.,
and Gartler, S. M. (1999). The DNMT3B DNA methyltransferase gene is mutated in the
ICF immunodeficiency syndrome. Proc Natl Acad Sci U S A 96, 14412-14417.
Heinzel, F. P., Sadick, M. D., and Locksley, R. M. (1988). Leishmania major: analysis of
lymphocyte and macrophage cellular phenotypes during infection of susceptible and
resistant mice. Exp Parasitol 65, 258-268.
Himmelrich, H., Parra-Lopez,
C., Tacchini-Cottier,
F., Louis, J. A., and Launois, P.
(1998). The IL-4 rapidly produced in BALB/c mice after infection with Leishmania
177
major down-regulates IL-12 receptor beta 2-chain expression on CD4+ T cells resulting
in a state of unresponsiveness to IL-12. J Immunol 161, 6156-6163.
Ho, I. C., Hodge, M. R., Rooney, J. W., and Glimcher, L. H. (1996). The proto-oncogene
c-maf is responsible for tissue-specific expression of interleukin-4. Cell 85, 973-983.
Hutchins, A. S., Artis, D., Hendrich, B. D., Bird, A. P., Scott, P., and Reiner, S. L.
(2005). Cutting edge: a critical role for gene silencing in preventing excessive type 1
immunity. J Immunol 175, 5606-5610.
Jeltsch, A. (2002). Beyond Watson and Crick: DNA methylation and molecular
enzymology of DNA methyltransferases. Chembiochem 3, 274-293.
Jones, P. L., Veenstra, G. J., Wade, P. A., Vermaak, D., Kass, S. U., Landsberger, N.,
Strouboulis, J., and Wolffe, A. P. (1998). Methylated DNA and MeCP2 recruit histone
deacetylase to repress transcription. Nat Genet 19, 187-191.
Lee, D. U., Agarwal, S., and Rao, A. (2002). Th2 lineage commitment and efficient IL-4
production involves extended demethylation of the IL-4 gene. Immunity 16, 649-660.
Lee, H. J., Takemoto, N., Kurata, H., Kamogawa, Y., Miyatake, S., O'Garra, A., and
Arai, N. (2000). GATA-3 induces T helper cell type 2 (Th2) cytokine expression and
chromatin remodeling in committed Thl cells. J Exp Med 192, 105-115.
Lehnertz, B., Ueda, Y., Derijck, A. A., Braunschweig, U., Perez-Burgos, L., Kubicek, S.,
Chen, T., Li, E., Jenuwein, T., and Peters, A. H. (2003). Suv39h-mediated histone H3
lysine 9 methylation directs DNA methylation to major satellite repeats at pericentric
heterochromatin. Curr Biol 13, 1192-1200.
Leonhardt, H., Page, A. W., Weier, H. U., and Bestor, T. H. (1992). A targeting sequence
directs DNA methyltransferase to sites of DNA replication in mammalian nuclei. Cell 71,
865-873.
Martins, G. A., Hutchins, A. S., and Reiner, S. L. (2005). Transcriptional activators of
helper T cell fate are required for establishment but not maintenance of signature
cytokine expression. J Immunol 175, 5981-5985.
Mullen, A. C., Hutchins, A. S., High, F. A., Lee, H. W., Sykes, K. J., Chodosh, L. A., and
Reiner, S. L. (2002). Hlx is induced by and genetically interacts with T-bet to promote
heritable T(H)1 gene induction. Nat Immunol 3, 652-658.
Murphy, K. M., and Reiner, S. L. (2002). The lineage decisions of helper T cells. Nat
Rev Immunol 2,933-944.
Nan, X., Ng, H. H., Johnson, C. A., Laherty, C. D., Turner, B. M., Eisenman, R. N., and
Bird, A. (1998). Transcriptional repression by the methyl-CpG-binding protein MeCP2
involves a histone deacetylase complex. Nature 393, 386-389.
178
Okano, M., Bell, D. W., Haber, D. A., and Li, E. (1999). DNA methyltransferases
Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development.
Cell 99, 247-257.
Okano, M., Xie, S., and Li, E. (1998). Cloning and characterization of a family of novel
mammalian DNA (cytosine-5) methyltransferases. Nat Genet 19, 219-220.
Ouyang, W., Ranganath, S. H., Weindel, K., Bhattacharya, D., Murphy, T. L., Sha, W.
C., and Murphy, K. M. (1998). Inhibition of Thl development mediated by GATA-3
through an IL-4-independent mechanism. Immunity 9, 745-755.
Sadick, M. D., Locksley, R. M., Tubbs, C., and Raff, H. V. (1986). Murine cutaneous
leishmaniasis: resistance correlates with the capacity to generate interferon-gamma in
response to Leishmania antigens in vitro. J Immunol 136, 655-661.
Weisenberger, D. J., Velicescu, M., Preciado-Lopez,
M. A., Gonzales, F. A., Tsai, Y. C.,
Liang, G., and Jones, P. A. (2002). Identification and characterization of alternatively
spliced variants of DNA methyltransferase 3a in mammalian cells. Gene 298, 91-99.
Winders, B. R., Schwartz, R. H., and Bruniquel, D. (2004). A distinct region of the
murine IFN-gamrma promoter is hypomethylated from early T cell development through
mature naive and Th cell differentiation, but is hypermethylated in Th2 cells. J Immunol
173, 7377-7384.
Xie, S., Wang, Z., Okano, M., Nogami, M., Li, Y., He, W. W., Okumura, K., and Li, E.
(1999). Cloning, expression and chromosome locations of the human DNMT3 gene
family. Gene 236, 87-95.
179