36
37
38
39
implantation in the mouse via PPARd.
Genes Dev. 13, 1561–1574
McAdam, B.F. et al. (1999) Systemic
biosynthesis of prostacyclin by cyclooxygenase (COX)-2: the human pharmacology
of a selective inhibitor of COX-2. Proc. Natl.
Acad. Sci. U. S. A. 96, 272–277
Brock, T.G. et al. (1999) Arachidonic acid is
preferentially metabolized by cyclooxygenase-2 to prostacyclin and prostaglandin
E2. J. Biol. Chem. 274, 11660–11666
Murata, T.F. et al. (1997) Altered pain perception and inflammatory response in
mice lacking prostacyclin receptor. Nature
388, 678–682
Forman, B.M. et al. (1997) Hypolipidemic
40
41
42
43
drugs, polyunsaturated fatty acids, and
eicosanoids are ligands for peroxisome
proliferator-activated receptors a and d.
Proc. Natl. Acad. Sci. U. S. A. 94, 4312–4317
Berger, J. et al. (1999) Novel PPARg and
PPARd ligands produce distinct biological
effects. J. Biol. Chem. 274, 6718–6725
Meade, E.A. et al. (1999) Peroxisome proliferators
enhance
cyclooxygenase-2
expression in epithelial cells. J. Biol. Chem.
274, 8328–8334
He, T.C. et al. (1999) PPARd is an APCregulated target of nonsteroidal antiinflammatory drugs. Cell 99, 335–345
Bonventre, J. et al. (1997) Reduced fertility
and postischaemic brain injury in mice
DNA Methylation and Silencing of
Gene Expression
John Newell-Price, Adrian J.L. Clark and Peter King
DNA methylation is associated with the silencing of gene expression. The
predominant mechanism involves the methylation of DNA and the subsequent recruitment of binding proteins that preferentially recognize
methylated DNA. In turn, these proteins associate with histone deacetylase and chromatin remodelling complexes to cause the stabilization of
condensed chromatin. Recent studies have indicated that the opposite
might also hold; namely, that targeting of methylation might depend on
altered chromatin structure. The family of methyltransferases and
methyl-binding proteins is expanding and becoming better characterized.
This review will focus on the mechanisms of methylation-associated
silencing of gene expression.
It is widely recognized that DNA methylation is associated with condensed
nuclease-resistant heterochromatin and
silencing of gene expression1,2. It plays
important roles in repressing the
expression of tumour suppressor genes
in cancer3, X chromosome inactivation4,
parental imprinting5,6 and is essential
for development7. The density of
methylation is important because a
weak promoter can be silenced by only
a few methylated CpGs, whereas a
J. Newell-Price, A.J.L. Clark and P. King are
at the Department of Endocrinology, St
Bartholomew’s and Royal London School of
Medicine and Dentistry, West Smithfield,
London, UK EC1A 7BE. Tel: 144 207 601
8343, Fax: 144 207 601 8505, e-mail: j.d.c.
newellprice@mds.qmw.ac.uk
142
higher density of methylation is
required to repress a strong promoter8.
Although distant methylated sequences
can contribute to repression9, the
repression is greater if the promoter
itself is methylated10.
How does the methylated DNA interfere with gene transcription? Within
the past two years, significant advances
have been made towards a better
understanding of the mechanisms
whereby this transcriptional repression
is mediated, with particular emphasis
on the role of binding proteins that
preferentially recognize methylated
DNA and recruit chromatin-remodelling factors. The recent characterization of new members of the methyltransferase family, the enzymes that
44
45
46
47
deficient in cytosolic phospholipase A2.
Nature 390, 622–625
Uozumi, N. et al. (1997) Role of cytosolic
phospholipase A2 in allergic response and
parturition. Nature 390, 618–622
Nguyen, M.T. et al. (1997) The
prostaglandin receptor EP4 triggers
remodelling of the cardiovascular system
at birth. Nature 390, 78–81
Sugimoto, Y. et al. (1997) Failure of parturition in mice lacking prostaglandin F
receptor. Science 277, 681–683
Thomas, D.W. et al. (1998) Coagulation
defects and altered hemodynamic responses
in mice lacking receptors for thromboxane
A2. J. Clin. Invest. 103, 1994–2001
methylate DNA, has provided insights
into how the pattern of methylation is
established in the genome. We will
focus this review on these developments to give an overview of the means
whereby DNA methylation might establish or enhance the silencing of gene
expression.
• Methylation Patterns
In mammals, methylation of the 59position of cytosine residues is a
reversible covalent modification of
DNA in the palindromic sequence 59CpG-39 (and occasionally 59-CpNpG-39)
with the methyl groups projecting into
the major groove. Adult methylation
patterns are reproduced at each round
of cell division, but the mechanisms
whereby the patterns are established
during development are far from clear.
What is known is that, shortly after
fertilization in the preimplantation
embryo, the methylation pattern of the
gametes is erased, with a dramatic
decrease in the total level of methylated
DNA. After implantation, a wave of de
novo methylation establishes a new pattern in which the vast majority of CpGs
are methylated11. A widely accepted
view is that during development nonCpG island promoters of various genes
undergo transcriptional activation and
demethylation at specific sequences in
a tissue-specific fashion2,12. Approximately 60% of genes have promoters
embedded within dense regions of
CpGs, known as CpG islands13. CpG
islands are thought to escape de novo
methylation through the action of Sp1
binding to sites in or near these regions
of DNA (Refs 14,15), and also through
1043-2760/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved. PII: S1043-2760(00)00248-4
TEM Vol. 11, No. 4, 2000
the action of an embryo-specific factor16. Indeed, the introduction of an in
vitro methylated CpG island into
embryonal cells results in its rapid complete demethylation17. Thus, in contrast
to the vast majority of CpGs in the
genome, which are methylated, CpG
islands are thought usually to be
unmethylated in all tissues. However,
the CpG islands of the testis-specific
histone gene H2b and the germ linespecific family of MAGE genes in the rat
and human respectively, are methylated in somatic tissues18–20. We have
also shown that the highly tissue-specific ‘pituitary’ promoter of the somatically expressed human proopiomelanocortin gene, POMC, which lies
within a CpG island, is methylated in
normal non-expressing tissue and that
methylation in vitro silences expression
(J. Newell-Price et al., unpublished).
This suggests that methylation might
be playing an important role in regulating tissue-specific expression, even for
certain genes whose promoters are
within CpG islands.
• DNA Methyltransferases
The question remains as to how these
patterns are initially established. The
process during development involves de
novo methylation, maintenance methylation and demethylation. Methyltransferases have been cloned (Table 1) and
partly characterized, but determination
of the process of demethylation has
remained more elusive.
Maintenance Methylation: DNA
Methyltransferase 1
The originally characterized DNA
methyltransferase (Dnmt1) is a maintenance methylase with preference in
vivo for hemimethylated DNA (Ref. 21).
This enzyme is recruited to replicating
DNA to reproduce the methylation pattern of the parental strands in the
daughter strands22. Mice deficient for
Dnmt1 die in mid-gestation, with significantly reduced levels of DNA methylation7. These mice also exhibit biallelic expression of imprinted genes
and ectopic X chromosome activation,
indicating the importance of methylation for these processes. However,
embryonic stem (ES) cells from these
TEM Vol. 11, No. 4, 2000
Table 1. DNA methyltransferases
Methyltransferase
Methylase
function
Main expression Disruption
pattern
in mice
Dnmt1
Dnmt3a
Dnmt3b
Maintenance
De novo
De novo
Adult/embryo
Embryo
Embryo
Mutations
in humans
Die in utero
–
Die at 4 weeks
–
Die in utero
ICF (Box 1)
Abbreviations: ICF, immunodeficiency, centromeric instability and facial anomalies syndrome.
mice are viable and capable of de novo
methylation, suggesting the existence
of independent de novo methylases23.
De novo Methylation: DNA
Methyltransferases 3a and 3b
Dnmt3a and Dnmt3b are two such
methyltransferases with de novo methyltransferase activity24. Recombinant
Dnmt3a and Dnmt3b are able to methylate CpG dinucleotides of unmethylated
and hemimethylated DNA in vitro and
the genes are expressed at high levels in
ES cells but at low levels in adult somatic
tissues25. Dnmt3a2/2 or Dnmt3b2/2 ES
cells are capable of de novo methylation
of introduced proviral DNA, and
Dnmt3a2/2 and Dnmt3b2/2 double
mutants exhibit no de novo methylase
activity, indicating that, together,
Dnmt3a and Dnmt3b are essential for de
novo methylation and that redundancy
of action exists. However, Dnmt3a2/2 or
Dnmt3b2/2 mice exhibit different phenotypes: the former survive to term but are
runted and die at four weeks of age,
whereas the Dnmt3b2/2 mice die in mid
gestation. Dnmt3b2/2 embryos and ES
cells exhibit demethylation of centromeric minor satellite repeats,
whereas Dnmt3a2/2 mice and ES cells
do not. In contrast, double-mutant mice
fail to develop beyond gastrulation.
Thus, these enzymes exhibit overlapping
but distinct functions.
Recently, compound heterozygous
mutations of human DNMT3B affecting
the catalytic domain of the enzyme have
been identified in immunodeficiency,
centromeric instability and facial
anomalies (ICF) syndrome25,26 (Box 1).
Lymphocytes from these patients exhibit demethylated centromeric regions
similar to that seen in Dnmt3b2/2 cells,
indicating that methylation by this
enzyme is essential for the organization
and stabilization of a specific type of
heterochromatin. However, in other
aspects, the phenotypes differ, with no
embryonic lethality in humans. The reason for this discrepancy is unclear, but
possibly relates to functional regions
outside the catalytic domain being present in the individuals with mutations
but absent from the mice because the
gene is deleted. Moreover, it is unclear
why this enzyme appears to be targeted
to particular regions of DNA, although
this might relate to the primary sequence of DNA or, alternatively, it might
be the result of altered chromatin structure (see below).
• Mechanisms of Silencing
How is the silencing achieved? Because
the methyl groups project into the
Box 1. Immunodeficiency, centromeric instability and facial anomalies
(ICF) syndrome
Clinical features
Hypertelorism
Low-set ears
Epicanthal folds
Macroglossia
Reduced immunoglobulins
Succumb to infectious disease
before adulthood
Cytogenetic features
Chromosomal fusions
Extension of juxtacentromeric
chromatin
Genetic features
Mutations of DNMT3B
143
Table 2. Methyl domain-binding proteins
Binding partners/contacts
MBD
Function
Core components
Specific components
MBD1
MBD2a
MBD2b
MBD3
MBD4
MeCP2
Repression
Repression
Putative demethylase/repression
Repression
DNA repair
Repression
Non-HDAC1 deacetylase
HDAC1/2, RbAP46/48
?Mi2/NuRD: Mi2b, MTA-2
HDAC1/2, RbAP46/48
Mi2/NuRD: Mi2b, MTA-2
HDAC1/2, RbAP46/48
Sin3A, SAP30, SAP18
Abbreviation: MBD, methylcytosine-binding domain.
major groove of DNA, one means is the
direct interference of the binding of
specific transcription factors that have
methylated
CpG(s)
within
their
response elements27. Conversely, other
factors such as Sp1 are indifferent to
CpG methylation. However, this mechanism is less important than the recruitment of methyl-CpG-binding proteins
that mediate repression. Five proteins
with homologous methylcytosine-binding domains (MBD) have now been
cloned: MBD1–4 (Ref. 28) and MeCP2
(Refs 29,30) (Table 2). Of these, MBD4
is a thymidine glycosylase repair enzyme, and is not associated with transcriptional inactivation; it is likely to
play a role in limiting the mutagenicity
of methylcytosine31. The remaining
members are transcriptional repressors.
MeCP2, MBD1, MBD2 and MBD3 differ
in their DNA-binding characteristics
and the precise means by which they
exert repressive effects, although a common theme is the recruitment of corepressors and histone deacetylase, leading to the remodelling of chromatin.
Box 2. Rett syndrome
Incidence 1 in 10 000–15 000
X–linked mutations of MECP2
Males die at birth
Progressive neurodevelopmental
disorder
Common cause of mental retardation
in females
Normal development until 6–19
months of age
Gradual loss of speech/purposeful
hand use
Microcephaly, seizures, autism
and ataxia
144
MeCP2
MeCP2 is a global repressor that binds
a single symmetrically methylated CpG
flanked by at least 6 bp of non-specific
DNA sequence32 and exerts repressive
effects over distances of several hundred
bp (Ref. 10). More binding sites than
MeCP2 molecules exist in the genome
and MeCP2 localizes to methyl-rich
mouse major satellite DNA (Refs 10,33),
and is able to displace bound histone 1.
This, together with the fact that a high
salt concentration (.0.5 M) is required
for elution from nuclei, suggests that
MeCP2 is tightly associated with
methylated DNA, and is likely to be
associated with permanent repression.
Although the 80 amino acid MBD of
MeCP2 is sufficient to recognize the
methylated CpG dinucleotide32, the Cterminal domain also facilitates binding34. The solution structure of the
MBD determined by nuclear magnetic
resonance (NMR) spectroscopy has
shown it to be an asymmetrical wedge
shape, and this implies, rather surprisingly, that the recognition does not utilize the symmetry of the methylated
CpG dinucleotide for binding35. The
transcriptional repression domain
(TRD) of MeCP2 resides in the C-terminal region. Co-transfection of a construct containing the TRD of MeCP2
fused to a GAL4 protein, with a reporter
construct containing a promoter with
GAL4-binding sites, resulted in the
silencing of the reporter construct. The
TRD immunoprecipitates the corepressor Sin3A, and histone deacetylase36,37.
In all, approximately 20% of the total
histone deacetylase activity in mammalian cells (or more in Xenopus cells)
is precipitated by MeCP2. Many transcriptional regulators exert their influence through histone acetylation
or deacetylation, with deacetylation
being associated with inactive heterochromatin38. The repression caused by
MeCP2 is largely relieved by the specific
deacetylase inhibitor trichostatin A
(TSA)39, indicating that much of the
repression is mediated by histone deacetylation and hence chromatin structure. Although MeCP2 deletion in ES
cells does not affect their viability, or
ability to differentiate, mice with a high
level of mutant chimeras for the deletion die at embryonic days (E) 8.5 to
12.5 (Ref. 40). The mice do not show the
bi-allelic expression of imprinted genes
or ectopic X chromosome activation
seen in Dnmt1-deficient mice. These
data underscored the importance of
MeCP2 for development and in mediating some of the repressive effects of
methylation, but suggested that other
MBDs are also likely to play important
roles. In humans, mutation of MECP2
on the X chromosome results in Rett
syndrome (Box 2), a progressive neurodevelopmental disorder, which is one of
the most common causes of mental
retardation in females41. Males survive
to birth but then die, whereas females
suffer loss of speech, autism, ataxia and
mental retardation, which develop
from six to 18 months of age. These
phenotypes differ from murine models
and might reflect the severity of disruption in the mouse chimeras.
MBD2
The more recently identified MBD2 (Ref.
28) is a component of the previously
TEM Vol. 11, No. 4, 2000
(a)
(b)
De novo
methylation
Methylated
DNA
Recruitment of
deacetylase
complex
Recruitment
of MBDs
De novo
methylation
Recruitment of
deacetylase
complex
Methylated
DNA
Deacetylated
condensed
chromatin
Recruitment of MBDs
and chromatin remodelling
complexes and stabilization
of condensed chain
trends in Endocrinology and Metabolism
Figure 1. Interplay between methylation and methylation patterns. DNA is depicted as a black line wrapped around a histone core. (a) De novo
methylation targeting histone deacetylation. (b) Chromatin-determining methylation patterns. Abbreviation: MBD, methylcytosine-binding
domain. Key: blue triangles, MBD proteins; green circles, deacetylase and remodelling complex; red circles, symmetrically methylated cytosines;
pink hexagons, de novo methylase.
identified MeCP1 complex42, and has
preference for methylated DNA. The
MeCP1 complex requires the presence
of multiple CpG dinucleotides for binding, and compared with MeCP2 has a
weaker association with methylated
DNA. MBD2 is able to immunoprecipitate more than 20% of the histone
deacetylase activity of HeLa cells,
which lack MeCP2. The components of
this complex include histone deacetylases HDAC1 and HDAC2, and the histone-binding proteins RbAp46 and
RbAp48 (Ref. 43), and might also
involve the same chromatin remodelling factors that are associated with
MBD3 (see below). The mechanism of
repression mediated by MBD2 appears
to vary depending on the promoter
studied. GAL4–MBD2 can repress the
transcription of the genes encoding
human DNA polymerase b and b-actin
TEM Vol. 11, No. 4, 2000
through GAL4 sites, but only the
repression of polymerase b production
is relieved by TSA (Ref. 43). Hence, the
repression of promoters by MBD2 is
not always dependent on the recruitment and action of histone deacetylation. Moreover, compared with
MeCP2, the lower affinity of MeCP1
for methylated DNA suggests that it
might have transient repressive roles.
MBD2 exists in two forms, MBD2a
and MBD2b, corresponding to initiation of translation from the first or second methionine codon28. Intriguingly,
MBD2b had been shown to exhibit
active demethylase activity in mammalian cells with the direct removal of
the methyl group44. However, two independent laboratories have not been
able to reproduce these results in mammalian and Xenopus systems43,45. It
might be that crucial cofactors were not
present in the studies on Xenopus systems, but current evidence favours a
transcriptional silencing role rather
than demethylation. Although other
systems with demethylase activity have
been reported46, our knowledge of this
process remains in its infancy.
MBD3
MBD3 shares sequence homology with
MBD2 outside the methyl-binding domain, but as assessed by conventional
purification methods, it exists in a complex distinct from MBD2. Mammalian
and Xenopus MBD3 forms part of the
repressor complex Mi2–NuRD (nucleosome remodelling histone deacetylase
complex)47,48. This contains the Mi2b
protein, which is a member of the
SW12/SNF superfamily of ATPases that
interferes with histone–DNA interactions and allows access of deacetylase
145
recruitment of the complex to methylated DNA (Ref. 47). Whether this is true
in vivo remains to be established, but it
might be that both MBD2 and MBD3
exert repressive effects through interaction with the NuRD complex.
microinjection of a methylated Xenopus heat shock protein (HSP) promoter
(which contained GAL4-binding sites)
that remained active for 10–12 h before
being silenced51. This silencing could
not be reversed by the strong activator
GAL4–VP16, indicating that once
assembled, condensed chromatin is
resistant to activation. Furthermore,
repressive chromatin has been shown
to spread from a focus of methylated
DNA (Ref. 52). Together, these data
suggest that chromatin assembly plays
a major role in the repressive effects of
CpG methylation, but that it is not the
whole story, because other promoters
are silenced in transient transfection
studies before completion of chromatin
assembly8,42. Interestingly, silenced
tumour suppressor genes of mammalian cells grown in culture, whose
CpG island promoters are methylated,
do not demonstrate reactivation after
treatment with TSA until the methylcytosine density is diminished by prior
incubation with the demethylating
agent 59-aza-29 deoxycytidine53. These
data suggest synergy of action for
repression between methylation and
deacetylation. Thus, although histone
deacetylation appears to be playing a
major role, some of the suppressive
effects of methylation are likely to be
mediated by mechanisms other than
chromatin condensation; for example,
by MBDs and recruited corepressors
interfering directly with the transcriptional machinery.
• DNA Methylation and Histone
Deacetylation: Layers of
Repression
Are all the effects of methylation mediated by effects on chromatin and
histone deacetylation? From the discussion above, it is apparent that not
all the repressive effects of MBD2 are
reversed by TSA, suggesting non-chromatin effects for silencing. However,
microinjection of a naked methylated
thymidine kinase gene promoter into
cultured mammalian cells resulted in
transcriptional activation for several
hours, whereas, when it was injected as
chromatin, there was immediate
repression50. Furthermore, time-dependent chromatin assembly was seen after
• Is Methylation Targeting
Chromatin or is Chromatin
Targeting Methylation?
De novo Methylation
Dnmt3 recognizes the centromeric
regions of the genome, and these are
undermethylated in ICF syndrome.
How these regions are targeted is
unknown. On the one hand, specific
sequences of DNA might preferentially
recruit this enzyme during development and allow de novo methylation,
with subsequent recruitment of MBDs,
histone remodelling complexes and the
formation of heterochromatin (Fig. 1a).
Alternatively, these regions of the
genome might exist in a chromatin
configuration favourable for de novo
Dnmt1 recruiting deacetylase
to the replication fork
Dnmt1 recruited to
deacetylated histones
trends in Endocrinology and Metabolism
Figure 2. The interplay between Dnmt1 and histone deacetylation. In the upper strand, Dnmt1
methylates the daughter strands after replication and recruits deacetylated histones to the nucleosome. Alternatively, Dnmt1 might require deacetylated histones for efficient maintenance
methylation of DNA. Abbreviations: Dnmt1, DNA methyltransferase 1; MBD, methylcytosinebinding domain. Key: blue triangles, MBD proteins; pink ovals, Dnmt1; green circles, deactylase
and remodelling complex; red circles, symmetrically methylated cytosines.
to histone tails. Mi2b is also the dermatomyositis autoantigen48,49. In addition to Mi2b, the Mi2–NuRD
complex includes RbAp46/RbAp48,
HDAC1/2 and MTA2 – a protein related
to Mta1 (metastasis-associated protein). This latter protein is important in
modulating the effects of the core deacetylase complex. Mammalian MBD3
has been shown to have equal preference for methylated and unmethylated
DNA in some assays28,47, whereas in
other gel shift experiments, mammalian and Xenopus MBD3 exhibits
preference for methylated DNA (Ref.
48). Moreover, in the mammalian
Mi2–NuRD complex, the predominant
form of MBD3 is the splice variant
MBD3b, which lacks a full methyl-cytosine-binding domain. Thus, it remains
to be fully established whether MBD3 is
recruiting the Mi2–NuRD complex to
methylated DNA. Interestingly, MBD2
can interact with NuRD in gel shift
assays and therefore might cause
146
TEM Vol. 11, No. 4, 2000
methylation, which then acts to stabilize the structure (Fig. 1b).
Maintenance Methylation
Dnmt1 has recently been shown to
associate directly with histone deacetylase54, providing a more direct link
between DNA methylation and chromatin structure. This suggests that the
type of chromatin might be essential in
maintaining methylation patterns.
With maintenance methylation, until
recently the situation was clear: Dnmt1
methylates hemimethylated DNA at the
replication fork, with the subsequent
recruitment of MBDs, followed by histone deacetylation and the condensation of chromatin. However, it is now
apparent that either Dnmt1 ensures
that nucleosomes are assembled from
deacetylated histones with further stabilization through the action of MBDs
and associated proteins such as Mi2
(Fig. 2, upper replicating DNA strand),
or that Dnmt1 requires chromatin to be
deacetylated before it can effectively
methylate DNA (Fig. 2, lower replicating strand).
• Conclusions and Perspectives
Our understanding of the interplay
between DNA methylation patterns and
chromatin structure is becoming more
integrated, with the two seemingly ever
more dependent on each other. Not all
of the repressive effects of methylation
pertain to histone deacetylation and
chromatin condensation, although
these appear to be the most important.
Only very recently have human disease
states (other than cancer) been identified that are associated with perturbations of these systems, and it seems
likely that more will follow. It remains
to be established whether de novo
methylases are associated directly with
deacetylase activity, and the roles of
MBD2 and MBD3 will be more fully
delineated by knockout studies in mice.
References
1 Keshet, I. et al. (1986) DNA methylation
affects the formation of active chromatin.
Cell 44, 535–543
2 Bird, A. (1992) The essentials of DNA
methylation. Cell 70, 5–8
3 Laird, P.W. and Jaenisch, R. (1996) The
role of DNA methylation in cancer genetics
TEM Vol. 11, No. 4, 2000
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
and epigenetics. Annu. Rev. Genet. 30,
441–464
Jaenisch, R. et al. (1922) Mammalian X
chromosome inactivation. Novartis Found.
Symp. 214, 200–209
Bartolomei, M.S. and Tilghman, S.M.
(1997) Genomic imprinting in mammals.
Annu. Rev. Genet. 31, 493–525
Li, E. et al. (1993) Role for DNA methylation in genomic imprinting. Nature 366,
362–365
Li, E. et al. (1992) Targeted mutation of the
DNA methyltransferase gene results in
embryonic lethality. Cell 69, 915–926
Boyes, J. and Bird, A. (1992) Repression of
genes by DNA methylation depends on
CpG density and promoter strength: evidence for involvement of a methyl-CpG
binding protein. EMBO J. 11, 327–333
Hug, M. et al. (1996) Transcriptional
repression by methylation: cooperativity
between a CpG cluster in the promoter and
remote CpG-rich regions. FEBS Lett. 379,
251–254
Nan, X. et al. (1997) MeCP2 is a transcriptional repressor with abundant binding
sites in genomic chromatin. Cell 88,
471–481
Jaenisch, R. and Jahner, D. (1984) Methylation, expression and chromosomal position of genes in mammals. Biochim.
Biophys. Acta 782, 1–9
Kafri, T. et al. (1992) Developmental pattern of gene-specific DNA methylation in
the mouse embryo and germ line. Genes
Dev. 6, 705–714
Antequera, F. and Bird, A. (1993) Number
of CpG islands and genes in human and
mouse. Proc. Natl. Acad. Sci. U. S. A. 90,
11995–11999
Brandeis, M. et al. (1994) Sp1 elements protect a CpG island from de novo methylation. Nature 371, 435–438
Macleod, D. et al. (1994) Sp1 sites in the
mouse aprt gene promoter are required to
prevent methylation of the CpG island.
Genes Dev. 8, 2282–2292
Siegfried, Z. et al. (1999) DNA methylation
represses transcription in vivo. Nat. Genet.
22, 203–206
Frank, D. et al. (1991) Demethylation of
CpG islands in embryonic cells. Nature 351,
239–241
Choi, Y.C. and Chae, C.B. (1991) DNA
hypomethylation and germ cell-specific
expression of testis-specific H2B histone
gene. J. Biol. Chem. 266, 20504–20511
Choi, Y.C. et al. (1996) Molecular cloning
of mouse somatic and testis-specific
H2B histone genes containing a methylated CpG island. DNA Cell Biol. 15,
495–504
De Smet, C. et al. (1999) DNA methylation
is the primary silencing mechanism for a
set of germ line-specific and tumourspecific genes with a CpG rich promoter.
Mol. Cell. Biol. 19, 7327–7335
Bestor, T.H. (1988) Cloning of a mammalian DNA methyltransferase. Gene 74,
9–12
Leonhardt, H. et al. (1992) A targeting
sequence directs DNA methyltransferase to
sites of DNA replication in mammalian
nuclei. Cell 71, 865–873
Lei, H. et al. (1996) De novo DNA cytosine
methyltransferase activities in mouse
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
embryonic stem cells. Development 122,
3195–3205
Okano, M. et al. (1998) Cloning and
characterization of a family of novel
mammalian DNA (cytosine-5) methyltransferases. Nat. Genet. 19, 219–220
Okano, M. et al. (1999) DNA methyltransferases Dnmt3a and Dnmt3b are essential
for de novo methylation and mammalian
development. Cell 99, 247–257
Xu, G.L. et al. (1999) Chromosome instability and immunodeficiency syndrome
caused by mutations in a DNA methyltransferase gene. Nature 402, 187–191
Tate, P.H. and Bird, A.P. (1993) Effects of
DNA methylation on DNA-binding proteins
and gene expression. Curr. Opin. Genet.
Dev. 3, 226–231
Hendrich, B. and Bird, A. (1998) Identification and characterization of a family of
mammalian methyl-CpG binding proteins.
Mol. Cell. Biol. 18, 6538–6547
Meehan, R.R. et al. (1989) Identification of
a mammalian protein that binds specifically to DNA containing methylated CpGs.
Cell 58, 499–507
Meehan, R.R. et al. (1992) Characterization
of MeCP2, a vertebrate DNA binding protein with affinity for methylated DNA.
Nucleic Acids Res. 20, 5085–5092
Hendrich, B. et al. (1999) The thymidine
glycosylase MBD4 can bind to the product
of deamination at methylated CpG sites.
Nature 401, 301–304
Nan, X. et al. (1993) Dissection of the
methyl-CpG binding domain from the
chromosomal protein MeCP2. Nucleic
Acids Res. 21, 4886–4892
Nan, X. et al. (1996) DNA methylation specifies chromosomal localization of MeCP2.
Mol. Cell. Biol. 16, 414–421
Chandler, S.P. et al. (1999) The methyl-CpG
binding transcriptional repressor MeCP2
stably associates with nucleosomal DNA.
Biochemistry 38, 7008–7018
Wakefield, R.I. et al. (1999) The solution
structure of the domain from MeCP2 that
binds to methylated DNA. J. Mol. Biol. 291,
1055–1065
Nan, X. et al. (1998) Transcriptional repression by the methyl-CpG-binding protein
MeCP2 involves a histone deacetylase complex. Nature 393, 386–389
Jones, P.L. et al. (1998) Methylated DNA
and MeCP2 recruit deacetylase and repress
transcription. Nat. Genet. 19, 187–191
Struhl, K. (1998) Histone acetylation and
transcriptional regulatory mechanisms.
Genes Dev. 12, 599–606
Yoshida, M. et al. (1995) Trichostatin A and
trapoxin: novel chemical probes for the
role of histone acetylation in chromatin
structure and function. BioEssays 17,
423–430
Tate, P. et al. (1996) The methyl-CpG binding protein MeCP2 is essential for embryonic development in the mouse. Nat. Genet.
12, 205–208
Amir, R.E. et al. (1999) Rett syndrome is
caused by mutations in X-linked MECP2,
encoding methyl-CpG-binding protein 2.
Nat. Genet. 23, 185–188
Boyes, J. and Bird, A. (1991) DNA methylation inhibits transcription indirectly via
a methyl-CpG binding protein. Cell 64,
1123–1134
147
43 Ng, H.H. et al. (1999) MBD2 is a transcriptional repressor belonging to the MeCP1
histone deacetylase complex. Nat. Genet.
23, 58–61
44 Bhattacharya, S.K. et al. (1999) A mammalian protein with specific demethylase
activity for mCpG DNA. Nature 397,
579–583
45 Wade, P.A. et al. (1998) A multiple subunit
Mi-2 histone deacetylase from Xenopus
laevis cofractionates with an associated
Snf2 superfamily ATPase. Curr. Biol. 8,
843–846
46 Weiss, A. et al. (1996) DNA demethylation
in vitro: involvement of RNA. Cell 86,
709–718
47 Zhang, Y. et al. (1999) Analysis of the NuRD
subunits reveals a histone deacetylase core
complex and a connection with DNA
methylation. Genes Dev. 13, 1924–1935
48 Wade, P.A. et al. (1999) Mi-2 complex
couples DNA methylation to chromatin
remodelling and histone deacetylation.
Nat. Genet. 23, 62–66
49 Zhang, Y. et al. (1998) The dermatomyositis-specific autoantigen Mi2 is a
component of a complex containing histone deacetylase and nucleosome remodeling activities. Cell 95, 279–289
50 Buschhausen, G. et al. (1987) Chromatin
structure is required to block transcription
of the methylated herpes simplex virus
thymidine kinase gene. Proc. Natl. Acad.
Sci. U. S. A. 84, 1177–1181
51 Kass, S.U. et al. (1997) DNA methylation
directs a time-dependent repression of
transcription initiation. Curr. Biol. 7,
157–165
52 Kass, S.U. et al. (1993) Inactive chromatin
spreads from a focus of methylation. Mol.
Cell. Biol. 13, 7372–7379
53 Cameron, E.E. et al. (1999) Synergy of
demethylation and histone deacetylase
inhibition in the re-expression of genes
silenced in cancer. Nat. Genet. 21, 103–107
54 Fuks, F. et al. (2000) DNA methyltransferase DnmT1 associates with histone
deacetylase activity. Nat. Genet. 24, 88–91
• Note Added in Proof
The suppressive effects of MDB1 have
recently been shown to be reversed by
treatment with TSA, suggesting that,
similarly to the other repressive MBDs,
the effects are mediated by recruitment
of histone deacetylation55. However,
unlike MBD2 and MeCP2, MBD1 is not
depleted by immunoprecipitation with
antibodies to HDAC1, implying that it
recruits a deacetylase distinct from the
core complex HDAC1/2.
55 Ng, H.H. et al. (2000) Active repression
of methylated genes by the chromosomal
protein MBD1. Mol. Cell. Biol. 20,
1394–1406
CONFERENCES
Digging the Genome for Diabetes
Mellitus: The 2nd ADA Research
Symposium on the Genetics of
Diabetes, San Jose, CA, USA,
17–19 October 1999
Michael A. Pani and Klaus Badenhoop
The diabetes genetics community recently met in California to discuss the
impact that the latest findings in susceptibility mapping might have on
our understanding of the pathophysiology of diabetes mellitus and its
long-term complications. Organized by Ralph DeFronzo (San Antonio,
TX) and Alan Permutt (St Louis, MO) the symposium covered the genetics of type 1 and type 2 diabetes, obesity, approaches to population selection and stratification, quantitative trait analysis, epidemiological tools
and technological advances in genotyping.
M.A. Pani and K. Badenhoop are at Med.
Klinik I, Klinikum der J.W. Goethe-Universität,
Theodor-Stern-Kai 7, D-60590 Frankfurt,
Germany. Tel: 149 69 6301 5839, Fax: 149
69 6301 6405, email: badenhoop@em.unifrankfurt.de
148
• Type 2 Diabetes: New Candidates
in Various Populations
Type 2 diabetes is a strong cofactor of
coronary heart disease and is a major
contributor to morbidity and mortality,
especially in the western world, where
it is highly prevalent. Although a strong
genetic component of this disease has
emerged, no gene has been so far discovered that confers a high risk for type
2 diabetes. Graeme Bell (Chicago, IL),
who had earlier identified a potential
susceptibility locus in a genome-wide
scan in Mexican Americans suffering
from type 2 diabetes1, has narrowed the
locus down to chromosomal region
2q37.3, containing 73 sequence-tagged
sites (STSs) in approximately 1.7
megabases. A particular single-nucleotide polymorphism (SNP), SNP43,
located in intron 3 of calpain 10
(CAPN10: a calcium-activated cysteinendopeptidase homologous to papain)
was found to be associated with higher
fasting glucose and lower sleeping
metabolic rate in type 2 diabetics. Supported by the fact that it also predicts
insulin resistance in nondiabetic individuals, Bell hypothesized that SNP43
in CAPN10 is the causal polymorphism
or is at least in very close linkage disequilibrium with the NIDDM1 locus. In
the course of saturation mapping in the
same population, Nancy Cox (Chicago,
1043-2760/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved. PII: S1043-2760(00)00246-0
TEM Vol. 11, No. 4, 2000