Functional Reassembly of Split Enzymes On‐Site: A Novel Approach for Highly Sequence‐Specific Targeted DNA Methylation

Abstract

In mammalian genomes, a significant fraction of the cytosine residues is methylated at the 5-position (5-methylcytosine), and this modified nucleobase is found in 5’-CG-3’ sequences (CpG sites). The methylation pattern of the genome changes during ontogenesis, depends on the tissue and can substantially differ in several diseases, notably cancer. We are only at the beginning of understanding the biological role of DNA methylation in higher organisms, but the emerging view is that methylation of the promoter region of a substantial fraction of genes leads to transcriptional inactivation (gene silencing). Recognition of the importance of DNA methylation in gene regulation raised the possibility of silencing selected genes by exogenous, targeted methylation of their promoters. Directing DNA methylation to predetermined sites, besides being a promising research tool for silencing genes of interest and studying DNA methylation in higher eukaryotes, could lead to therapeutic applications in diseases characterised by aberrant expression of one or a small number of genes. A potential advantage of DNA methylation-mediated gene silencing is that the de novo established methylation pattern is stably propagated by maintenance methylation through cycles of semiconservative replication. The silencing mechanism is likely to involve other epigenetic factors (e.g. , histone methyltransferases, histone deacetylases) recruited by the DNA methylation marks. DNA methylation is catalysed by DNA methyltransferases (MTases), which transfer the activated methyl group from the ubiquitous cofactor S-adenosyl-l-methionine (AdoMet or SAM) to their target nucleobase within short DNA recognition sequences ranging from two to eight base pairs (bps). Directing DNA MTases to a preselected recognition sequence (targeted DNA methylation) was pioneered by Xu and Bestor, who genetically fused the bacterial DNA (cytosine-C5) MTase M.SssI (recognition sequence 5’CG-3’) to a zinc finger protein (ZFP) that recognises a 9 bp DNA sequence with high specificity and serves as targeting domain. The idea was that the chimeric protein would bind to the DNA sequence specific for the targeting domain, and the DNA MTase would selectively methylate CpG sequences located close to the binding site of the fusion partner. Indeed, preferential methylation of a CpG sequence located in the vicinity of the ZFP binding site was observed in in vitro experiments. Later research demonstrated that this approach can, in principle, also be applied in vivo for targeting chromosomal as well as mitochondrial DNA. In the most extensive study so far, the catalytic domains of the murine DNA MTases Dnmt3a and Dnmt3b were fused to sequence-specific DNA-binding proteins. In transient cotransfection experiments on human cells, dense methylation of the targeted promoter regions and silencing of the targeted genes was demonstrated. In almost all such studies, ZFPs were used as targeting devices because customised ZFPs can now be engineered to bind with high affinity and sequence specificity to almost any DNA sequence. A key issue of targeted DNA methylation is specificity, that is, the difference between levels of exogenous methylation at the targeted site and nontargeted sites. Analysis of the methylation status of nontargeted sites revealed either methylation far from the target site, 8] or extensive methylation of regions flanking the binding site of the targeting domain. Although efficient gene silencing might require methylation of many CpG sites in a promoter, and thus the latter phenomenon could be an advantage in many cases, high-resolution analysis of the effect of DNA methylation would require a method suitable for the methylation of single CpG sites. Methylation of nontargeted sites is not surprising owing to the inherent affinity of DNA MTases for their recognition sequences (Scheme 1A). Nontargeted methylation limits the use of simple DNA MTase fusions as a research tool and would be a serious obstacle to therapeutic application. Therefore, novel strategies that allow highly sequence-specific targeted DNA methylation are of prime importance. One approach to improving targeting specificity employs DNA MTase variants with reduced DNA-binding affinity. Here, DNA binding of the mutant DNA MTase-ZFP fusions is dominated by the targeting ZFP domain’s improving the specificity. A recent paper by Nomura and Barbas describes another approach that appears to be a significant step towards the goal of methylating single CpG sites within whole genomes. The authors capitalised on previous observations of protein fragment complementa[a] Prof. Dr. E. Weinhold Institut f r Organische Chemie RWTH Aachen University Landoltweg 1, 52056 Aachen (Germany) Fax: (+49)241-80-92528 E-mail : elmar.weinhold@oc.rwth-aachen.de [b] Prof. A. Kiss Institute of Biochemistry Biological Research Centre of the Hungarian Academy of Sciences Temesv7ri krt. 62, 6726 Szeged (Hungary) Fax: (+36)62-433-506 E-mail : kissa@brc.hu