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