M.EcoRV is an alpha-adenine DNA methyltransferase. According to structure predictions, the enzyme consists of a catalytic domain, which has a structure similar to all other DNA-methyltransferases, and a smaller DNA-recognition domain. We have investigated this enzyme by random mutagenesis, using error-prone PCR, followed by selection for catalytically inactive mutants. 20 single mutants were identified that are completely inactive in vivo as His6- and GST-fusion proteins. 13 of them could be overexpressed and purified. All of these mutants are also inactive in vitro. 5 of the mutations are located near the putative binding site for a flipped adenine residue (C192R, D193G, E212G, W231R, N239H). All of these variants bind to DNA, demonstrating the importance of this region of the protein in catalysis. Only the W231R mutant could be purified with high yields. It binds to DNA and AdoMet and, thus, behaves like a bona fide active site mutant. According to the structure prediction Trp231 corresponds to Val121 in M.HhaI, which forms a hydrophobic contact to the flipped target cytosine. 4 of the remaining purified variants are located within a small region of the putative DNA-recognition domain (F115S, F117L, S121P, C122Y). F117L, S121P and C122Y are unable to bind to DNA, suggesting a critical role of this region in DNA binding. Taken together, these results are in good agreement with the structural model of M.EcoRV.
DNA methylation patterns of mammalian genomes are generated in gametogenesis and early embryonic development. Two de novo DNA methyltransferases, Dnmt3a and Dnmt3b, are responsible for the process. Both enzymes contain a long N-terminal regulatory region linked to a conserved C-terminal domain responsible for the catalytic activity. Although a PWWP domain in the N-terminal region has been shown to bind DNA in vitro, it is unclear how the DNA methyltransferases access their substrate in chromatin in vivo. We show here that the two proteins are associated with chromatin including mitotic chromosomes in mammalian cells, and the PWWP domain is essential for the chromatin targeting of the enzymes. The functional significance of PWWP-mediated chromatin targeting is suggested by the fact that a missense mutation in this domain of human DNMT3B causes immunodeficiency, centromeric heterochromatin instability, facial anomalies (ICF) syndrome, which is characterized by loss of methylation in satellite DNA, pericentromeric instability, and immunodeficiency. We demonstrate that the mutant protein completely loses its chromatin targeting capacity. Our data establish the PWWP domain as a novel chromatin/chromosome-targeting module and suggest that the PWWP-mediated chromatin association is essential for the function of the de novo methyltransferases during development. DNA methylation patterns of mammalian genomes are generated in gametogenesis and early embryonic development. Two de novo DNA methyltransferases, Dnmt3a and Dnmt3b, are responsible for the process. Both enzymes contain a long N-terminal regulatory region linked to a conserved C-terminal domain responsible for the catalytic activity. Although a PWWP domain in the N-terminal region has been shown to bind DNA in vitro, it is unclear how the DNA methyltransferases access their substrate in chromatin in vivo. We show here that the two proteins are associated with chromatin including mitotic chromosomes in mammalian cells, and the PWWP domain is essential for the chromatin targeting of the enzymes. The functional significance of PWWP-mediated chromatin targeting is suggested by the fact that a missense mutation in this domain of human DNMT3B causes immunodeficiency, centromeric heterochromatin instability, facial anomalies (ICF) syndrome, which is characterized by loss of methylation in satellite DNA, pericentromeric instability, and immunodeficiency. We demonstrate that the mutant protein completely loses its chromatin targeting capacity. Our data establish the PWWP domain as a novel chromatin/chromosome-targeting module and suggest that the PWWP-mediated chromatin association is essential for the function of the de novo methyltransferases during development.
Abstract DNA methyltransferases (DNMTs) are responsible for DNA methylation, an epigenetic modification involved in gene regulation. Families of conjugates of procainamide, an inhibitor of DNMT1, were conceived and produced by rapid synthetic pathways. Six compounds resulted in potent inhibitors of the murine catalytic Dnmt3A/3L complex and of human DNMT1, at least 50 times greater than that of the parent compounds. The inhibitors showed selectivity for C5 DNA methyltransferases. The cytotoxicity of the inhibitors was validated on two tumour cell lines (DU145 and HCT116) and correlated with the DNMT inhibitory potency. The inhibition potency of procainamide conjugated to phthalimide through alkyl linkers depended on the length of the linker; the dodecane linker was the best.
SETDB1 is a H3K9 methyltransferase involved in transcriptional silencing with a catalytic SET domain and a triple Tudor domain containing a methyl-lysine binding site. SGC Toronto previously solved the apo structure of the Tudor domain (PDB code 3DLM). Amplification of SETDB1 in over 15% lung adenocarcinoma correlates with high mRNA and protein levels and its depletion in SETDB1-amplified cells reduces cancer growth in cell culture and nude mice models, whereas its overexpression increases tumour invasiveness (Rodriguez-Paredes et al. Oncogene 2014, Shah et al. Epigenetic Chromatin 2014). Several histone methyltransferases are known to have non-catalytic functions that might be alternative targeting strategies. For instance, recognition of H3K9 methylation by the ankyrin repeat of the methyltransferase GLP is required for efficient establishment of H3K9 methylation (Liu et al. Genes Dev. 2015). No catalytic domain inhibitor of SETDB1 has been reported to date. The goal of this TEP is to enable the discovery of potent, selective compounds targeting the Tudor domain of SETDB1.
Lysine methylation of histones and non-histone substrates by the SET domain containing protein lysine methyltransferase (KMT) G9a/EHMT2 governs transcription contributing to apoptosis, aberrant cell growth, and pluripotency. The positioning of chromosomes within the nuclear three-dimensional space involves interactions between nuclear lamina (NL) and the lamina-associated domains (LAD). Contact of individual LADs with the NL are dependent upon H3K9me2 introduced by G9a. The mechanisms governing the recruitment of G9a to distinct subcellular sites, into chromatin or to LAD, is not known. The cyclin D1 gene product encodes the regulatory subunit of the holoenzyme that phosphorylates pRB and NRF1 thereby governing cell-cycle progression and mitochondrial metabolism. Herein, we show that cyclin D1 enhanced H3K9 dimethylation though direct association with G9a. Endogenous cyclin D1 was required for the recruitment of G9a to target genes in chromatin, for G9a-induced H3K9me2 of histones, and for NL-LAD interaction. The finding that cyclin D1 is required for recruitment of G9a to target genes in chromatin and for H3K9 dimethylation, identifies a novel mechanism coordinating protein methylation.