The level of genomic DNA methylation plays an important role in development and disease. In order to establish an experimental system for the functional analysis of genome-wide hypermethylation, we overexpressed the mouse de novo methyltransferase Dnmt3a in Drosophila melanogaster. These flies showed severe developmental defects that could be linked to reduced rates of cell cycle progression and irregular chromosome condensation. In addition, hypermethylated chromosomes revealed elevated rates of histone H3-K9 methylation and a more restricted pattern of H3-S10 phosphorylation. The developmental and chromosomal defects induced by DNA hypermethylation could be rescued by mutant alleles of the histone H3-K9 methyltransferase gene Su(var)3-9. This mutation also resulted in a significantly decreased level of genomic DNA methylation. Our results thus uncover the molecular consequences of genomic hypermethylation and demonstrate a mutual interaction between DNA methylation and histone methylation.Aberrant DNA methylation patterns are one of the most consistent hallmarks of cancer (31). Many of the epigenetic aberrations that occur during tumorigenesis can be related to ectopic hypermethylation (4). While the analysis by various experimental systems of reduced levels of DNA methylation has helped to define the role of DNA methylation, the functional consequences of DNA hypermethylation remain unclear. DNA methylation is catalyzed by a class of enzymes called DNA methyltransferases (7). During early mammalian embryogenesis, de novo methyltransferases catalyze the methylation of previously unmethylated cytosine residues. During later stages, DNA methylation patterns are copied postreplicatively from the parental strand to the daughter strand (29). This step requires a functionally specialized DNA methyltransferase that has been termed the maintenance methyltransferase. To date, there are four known mammalian DNA methyltransferase genes. Dnmt1 encodes an enzyme with maintenance methyltransferase activity (8,22), while the closely related genes Dnmt3a and Dnmt3b encode de novo methyltransferases (26,39,48). The function of the fourth gene, Dnmt2, remains enigmatic (16,49,65).Disruption of DNA methyltransferase activity proved that DNA methylation was essential for proper development of mice (35), frogs (59), and plants (18,54). The analysis of corresponding DNA hypomethylation phenotypes identified a role for DNA methylation in the regulation of gene activity (28, 59), control of foreign DNA elements (42, 63), and genome stability (14). Similar conclusions have been reached from the analysis of ICF (immunodeficiency, centromeric instability, facial anomalies) syndrome patients. These patients lack the majority of DNMT3B methyltransferase activity (23,48,64) and show striking chromosome aberrations that have been attributed to the demethylation and concomitant decondensation of pericentromeric satellite DNA (57). Together, these results suggest that DNA methylation might influence yet undefined aspects of higher-order chrom...

Epigenetic information is encoded by DNA methylation and by covalent modifications of histone tails. While defined epigenetic modification patterns have been frequently correlated with particular states of gene activity, very little is known about the integration level of epigenetic signals. Recent experiments have resulted in the characterization of several epigenetic adaptors that mediate interactions between distinct modifications. These adaptors include methyl-DNA binding proteins, chromatin remodelling enzymes and siRNAs. Complex interactions between epigenetic modifiers and adaptors provide the foundation for the stability of epigenetic inheritance. In addition, they also provide an explanation for the long-range effects of epigenetic mechanisms. We propose that a major aspect of epigenetic regulation lies in the modification of chromosome architecture and that local changes in gene expression would be secondary consequences. This view is consistent with many results from recent genomic analyses.