Laboratories

Research

Main Scientific Achievements

• We discovered the structural basis for sensing DNA modifications by modification-dependent endonucleases.
• We contributed structures of 6mA-sensing endonucleases of DpnI and discovered a novel, steric conflict-based readout mechanism that detects 6-methyladenine in DNA.
• We contributed structures of several families of 5-methylcytosine-, 5-hydroxymethylcytosine-, and glucosyl-5-hydroxymethylcytosine-dependent endonucleases and showed that they share a two-domain architecture with the PUA-superfamily or NEco modification sensor and PD-(D/E)XK or HNH catalytic domain.

Research Description

Our laboratory is interested in epigenetics from a biochemical perspective. We want to learn how methyl groups are specifically placed, read, and eventually removed. We take advantage of the fact that methylation is a biochemically ancient process so we can answer mechanistic questions using prokaryotic models. For more biological questions, we must work with mammalian proteins, cells, and animal models. Some of the key conclusions of our work are the following:

DNA adenine and cytosine methylation have opposite effects on the stability of double-stranded DNA. Their cellular functions in eukaryotes are broadly consistent with these biochemical tendencies, perhaps because evolution has built on the trends. In prokaryotic models, we have shown that apart from reliance on a hydrophobic effect, adenine and cytosine methylation are read quite differently. Adenine methylation is detected based on induced DNA deformation that is attributable to a methyl-methyl-clash (Mierzejewska et al. 2014), without nucleotide flipping. Cytosine hemi-methylation (or hemi-hydroxymethylation) is read through nucleotide flipping (Kisiala et al. 2018; Kazrani et al. 2014). Full cytosine methylation (in both DNA strands) is selected in the regular dsDNA context without nucleotide flipping (Slyvka et al. 2019). We have further identified a very limited toolkit for the detection of methylation that relies on a winged-helix modification sensor (Mierzejewska et al. 2014) for adenine methylation and an SRA domain sensor for cytosine hemi-(hydroxy)methylation, even in cases in which bioinformatic analysis suggested an unrelated domain for modification detection or was inconclusive (Kisiala et al. 2018; Kazrani et al. 2014). More recently, we discovered a novel domain (NEco) for cytosine methylation readout in both DNA strands that is unrelated to the previously implicated Zn²⁺ finger/MBD/MeCP2 domain family (Slyvka et al. 2019; Czapinska et al. 2018). We have also shown that PUA-superfamily domains other than SRA that were previously implicated in RNA binding are also naturally used for the binding of modified cytosines (Lutz et al. 2019).

In eukaryotes, active DNA demethylation can be reversed by up to three sequential hydroxylation reactions that convert 5mC into bases that resemble DNA damage. Such bases can then be excised and replaced by DNA repair machinery (Bochtler et al. 2017). We have shown that the hydroxylation system that was previously described in vertebrates is also present and functional in some invertebrates (Wojciechowski et al. 2014). Moreover, we have extensive (still unpublished) data that were obtained together with the groups of Tomasz Jurkowski and Tim Hore that suggest the strong sequence specificity of TET enzymes and crystal structures that explain this specificity. Finally, we have also worked on the base excision repair step of DNA demethylation and demonstrated cooperation (and mutual stimulation) between TDG and NEIL1, a bifunctional DNA glycosylase that was previously mostly implicated in the repair of oxidative DNA damage (Slyvka et al. 2017).

• Bibliography

1. Xu et al. Nat. Chem. 2020. doi: 10.1038/s41589-020-00675-5
2. Kisiala et al. NAR. 2020. doi: 10.1093/nar/gkaa403
3. Lutz et al. NAR. 2019. doi: 10.1093/nar/gkz755
4. Slyvka et al. NAR. 2019. doi: 10.1093/nar/gkz1017
5. Kisiala et al. NAR. 2018. doi: 10.1093/nar/gky781
6. Czapinska et al. NAR. 2018. doi: 10.1093/nar/gky731
7. Mierzejewska et al. NAR. 2014. doi: 1093/nar/gku546
8. Kazrani et al. NAR. 2014. doi: 10.1093/nar/gku186
9. Wojciechowski et al. PNAS. 2013. doi: 1073/pnas.1207986110
10. Siwek et al. NAR. 2012. doi: 10.1093/nar/gks428

Methodology

Much of our work involves classic protein biochemistry, combined with structural biology. In the past, this has meant primarily X-ray crystallography, but we are now also learning cryo-electron microscopy to routinely tackle larger protein complexes. In addition to our structural biology work, we have increasingly relied on various high-throughput sequencing methods for our research, often in combination with bisulfite sequencing to assess DNA modification status.

Selected Publications

• Reversal of nucleobase methylation by dioxygenases. Xu G-L, Bochtler M. Nat. Chem. Biol. 2020. doi: 10.1038/s41589-020-00675-5
• Arrhenius-law-governed homo- and heteroduplex dissociation. Bochtler M. Phys. Rev. E. 2020. doi: 10.1103/PhysRevE.101.032405
• Activity and structure of EcoKMcrA. Czapinska, H, Kowalska M, Zagorskaitė E, Manakova E, Slyvka A, Xu, Siksnys V, Sasnauskas G, Bochtler M. NAR. 2018, doi: 10.1093/nar/gky731
• Crystal structure of the 5hmC specific endonuclease PvuRts1I. Kazrani A A, Kowalska M, Czapinska H, Bochtler M. NAR. 2014. doi: 10.1093/nar/gku186
• CpG underrepresentation and the bacterial CpG-specific DNA methyltransferase M.MpeI. Wojciechowski M, Czapinska H, Bochtler M. PNAS. 2013. doi: 10.1073/pnas.1207986110

Collaborations

• Tomasz Jurkowski, Cardiff University, United Kingdom
• Tim Hore, Otaga University, New Zealand
• Guoliang Xu, Institute of Biochemistry, Chinese Academy of Sciences, China
• Jiemin Weng, Northeastern Normal University, China
• Virginijus Siksnys, Vilnius University, Lithuania
• Saulius Klimasauskas, Vilnius University, Lithuania
• Albert Jeltsch, Stuttgart University, Germany