Methylation at the 5 position of cytosine in DNA (5meC) is a key epigenetic mark in eukaryotes. Once introduced, 5meC can be maintained through DNA replication by the activity of ‘maintenance’ DNA methyltransferases (DNMTs). Despite their ancient origin, DNA methylation pathways differ widely across animals, such that 5meC is either confined to transcribed genes or lost altogether in several lineages. We used comparative epigenomics to investigate the evolution of DNA methylation. Although the model nematode Caenorhabditis elegans lacks DNA methylation, more basal nematodes retain cytosine DNA methylation, which is targeted to repeat loci. We found that DNA methylation coevolved with the DNA alkylation repair enzyme ALKB2 across eukaryotes. In addition, we found that DNMTs introduced the toxic lesion 3-methylcytosine into DNA both in vitro and in vivo. Alkylation damage is therefore intrinsically associated with DNMT activity, and this may promote the loss of DNA methylation in many species.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Additional information

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


  1. 1.

    Law, J. A. & Jacobsen, S. E. Establishing, maintaining and modifying DNA methylation patterns in plants and animals. Nat. Rev. Genet. 11, 204–220 (2010).

  2. 2.

    Holliday, R. Epigenetics: a historical overview. Epigenetics 1, 76–80 (2006).

  3. 3.

    Ponger, L. & Li, W. H. Evolutionary diversification of DNA methyltransferases in eukaryotic genomes. Mol. Biol. Evol. 22, 1119–1128 (2005).

  4. 4.

    Jurkowski, T. P. & Jeltsch, A. On the evolutionary origin of eukaryotic DNA methyltransferases and Dnmt2. PLoS ONE 6, e28104 (2011).

  5. 5.

    Zemach, A., McDaniel, I. E., Silva, P. & Zilberman, D. Genome-wide evolutionary analysis of eukaryotic DNA methylation. Science 328, 916–919 (2010).

  6. 6.

    Feng, S. et al. Conservation and divergence of methylation patterning in plants and animals. Proc. Natl. Acad. Sci. USA 107, 8689–8694 (2010).

  7. 7.

    Lyko, F. et al. The honey bee epigenomes: differential methylation of brain DNA in queens and workers. PLoS Biol. 8, e1000506 (2010).

  8. 8.

    Xiang, H. et al. Single base–resolution methylome of the silkworm reveals a sparse epigenomic map. Nat. Biotechnol. 28, 516–520 (2010).

  9. 9.

    Bewick, A. J., Vogel, K. J., Moore, A. J. & Schmitz, R. J. Evolution of DNA methylation across insects. Mol. Biol. Evol. 34, 654–665 (2017).

  10. 10.

    Raddatz, G. et al. Dnmt2-dependent methylomes lack defined DNA methylation patterns. Proc. Natl. Acad. Sci. USA 110, 8627–8631 (2013).

  11. 11.

    Goll, M. G. et al. Methylation of tRNAAsp by the DNA methyltransferase homolog Dnmt2. Science 311, 395–398 (2006).

  12. 12.

    Gao, F. et al. Differential DNA methylation in discrete developmental stages of the parasitic nematode Trichinella spiralis. Genome Biol. 13, R100 (2012).

  13. 13.

    Schiffer, P. H. et al. The genome of Romanomermis culicivorax: revealing fundamental changes in the core developmental genetic toolkit in Nematoda. BMC Genomics 14, 923 (2013).

  14. 14.

    Li, Z. et al. Distinct roles of DNMT1-dependent and DNMT1-independent methylation patterns in the genome of mouse embryonic stem cells. Genome Biol. 16, 115 (2015).

  15. 15.

    Huff, J. T. & Zilberman, D. Dnmt1-independent CG methylation contributes to nucleosome positioning in diverse eukaryotes. Cell 156, 1286–1297 (2014).

  16. 16.

    Falckenhayn, C. et al. Characterization of genome methylation patterns in the desert locust Schistocerca gregaria. J. Exp. Biol. 216, 1423–1429 (2013).

  17. 17.

    Kao, D. et al. The genome of the crustacean Parhyale hawaiensis, a model for animal development, regeneration, immunity and lignocellulose digestion. eLife 5, e20062 (2016).

  18. 18.

    Sarkies, P. et al. Ancient and novel small RNA pathways compensate for the loss of piRNAs in multiple independent nematode lineages. PLoS Biol. 13, e1002061 (2015).

  19. 19.

    Ougland, R., Rognes, T., Klungland, A. & Larsen, E. Non-homologous functions of the AlkB homologs. J. Mol. Cell Biol. 7, 494–504 (2015).

  20. 20.

    Sedgwick, B. Repairing DNA-methylation damage. Nat. Rev. Mol. Cell Biol. 5, 148–157 (2004).

  21. 21.

    Strepetkaitė, D. et al. Analysis of DNA methylation and hydroxymethylation in the genome of crustacean Daphnia pulex. Genes 7, 1 (2015).

  22. 22.

    Ringvoll, J. et al. Repair deficient mice reveal mABH2 as the primary oxidative demethylase for repairing 1meA and 3meC lesions in DNA. EMBO J. 25, 2189–2198 (2006).

  23. 23.

    Nay, S. L., Lee, D.-H., Bates, S. E. & O’Connor, T. R. Alkbh2 protects against lethality and mutation in primary mouse embryonic fibroblasts. DNA Repair 11, 502–510 (2012).

  24. 24.

    Gowher, H. et al. Mutational analysis of the catalytic domain of the murine Dnmt3a DNA-(cytosine C5)-methyltransferase. J. Mol. Biol. 357, 928–941 (2006).

  25. 25.

    Klimasauskas, S., Kumar, S., Roberts, R. J. & Cheng, X. HhaI methyltransferase flips its target base out of the DNA helix. Cell 76, 357–369 (1994).

  26. 26.

    Tsumura, A. et al. Maintenance of self-renewal ability of mouse embryonic stem cells in the absence of DNA methyltransferases Dnmt1, Dnmt3a and Dnmt3b. Genes Cells 11, 805–814 (2006).

  27. 27.

    Sved, J. & Bird, A. The expected equilibrium of the CpG dinucleotide in vertebrate genomes under a mutation model. Proc. Natl. Acad. Sci. USA 87, 4692–4696 (1990).

  28. 28.

    Alexandrov, L. B. et al. Clock-like mutational processes in human somatic cells. Nat. Genet. 47, 1402–1407 (2015).

  29. 29.

    Drabløs, F. et al. Alkylation damage in DNA and RNA—repair mechanisms and medical significance. DNA Repair 3, 1389–1407 (2004).

  30. 30.

    Furrer, A. & van Loon, B. Handling the 3-methylcytosine lesion by six human DNA polymerases members of the B-, X- and Y-families. Nucleic Acids Res. 42, 553–566 (2014).

  31. 31.

    Chastain, P. D. II et al. Abasic sites preferentially form at regions undergoing DNA replication. FASEB J. 24, 3674–3680 (2010).

  32. 32.

    Shrivastav, M., De Haro, L. P. & Nickoloff, J. A. Regulation of DNA double-strand break repair pathway choice. Cell Res. 18, 134–147 (2008).

  33. 33.

    Blaxter, M. L. et al. A molecular evolutionary framework for the phylum Nematoda. Nature 392, 71–75 (1998).

  34. 34.

    Quinlan, A. R. & Hall, I. M. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26, 841–842 (2010).

  35. 35.

    Emperle, M., Rajavelu, A., Reinhardt, R., Jurkowska, R. Z. & Jeltsch, A. Cooperative DNA binding and protein/DNA fiber formation increases the activity of the Dnmt3a DNA methyltransferase. J. Biol. Chem. 289, 29602–29613 (2014).

Download references


We thank H. Leitch and M. Borkowska for invaluable help with mouse ESC culture. We would like to thank M. Merkenschlager, L. Aragon, J. Sale and B. Lehner for helpful comments on the manuscript, M. Blaxter for advice on nematode genomics, and M. Berriman for access to the N. brasiliensis draft genome. P.S. is funded by an Imperial College Research Fellowship. Work in the Sarkies and Hajkova laboratories is funded by the Medical Research Council. P.H. is a recipient of the ERC CoG grant “dynamic modifications” and a member of the EMBO Young Investigator Programme. A.J. and M.E. are funded by DFG JE252/10. R.K.G. and A.J.B. are funded by Wellcome Trust grant 083620Z and Centre grant 203128/Z/16/Z. P.H.S. is funded by the ERC in a grant to Max Telford (ERC-2012-AdG 322790).

Author information

Author notes

  1. These authors contributed equally: Silvana Rošić, Rachel Amouroux and Cristina E. Requena.


  1. MRC London Institute of Medical Sciences, London, UK

    • Silvana Rošić
    • , Rachel Amouroux
    • , Cristina E. Requena
    • , Ana Gomes
    • , Toni Beltran
    • , Jayant K. Rane
    • , Sarah Linnett
    • , Petra Hajkova
    •  & Peter Sarkies
  2. Institute of Clinical Sciences, Imperial College London, London, UK

    • Silvana Rošić
    • , Rachel Amouroux
    • , Cristina E. Requena
    • , Ana Gomes
    • , Toni Beltran
    • , Jayant K. Rane
    • , Sarah Linnett
    • , Petra Hajkova
    •  & Peter Sarkies
  3. Institute of Biochemistry, Universität Stuttgart, Stuttgart, Germany

    • Max Emperle
    •  & Albert Jeltsch
  4. Department of Life Sciences, Imperial College London, London, UK

    • Murray E. Selkirk
  5. Department of Ecology and Evolution, University College London, London, UK

    • Philipp H. Schiffer
  6. School of Biological Sciences and Wellcome Trust Centre for Cell Matrix Research, FBMH, MAHSC, University of Manchester, Manchester, UK

    • Allison J. Bancroft
    •  & Richard K. Grencis


  1. Search for Silvana Rošić in:

  2. Search for Rachel Amouroux in:

  3. Search for Cristina E. Requena in:

  4. Search for Ana Gomes in:

  5. Search for Max Emperle in:

  6. Search for Toni Beltran in:

  7. Search for Jayant K. Rane in:

  8. Search for Sarah Linnett in:

  9. Search for Murray E. Selkirk in:

  10. Search for Philipp H. Schiffer in:

  11. Search for Allison J. Bancroft in:

  12. Search for Richard K. Grencis in:

  13. Search for Albert Jeltsch in:

  14. Search for Petra Hajkova in:

  15. Search for Peter Sarkies in:


P.S. and P.H. conceived the study. P.S., P.H. and A.J. designed the experiments. DNA extraction and bisulfite sequencing were carried out by S.R. and P.S. P.S. performed bioinformatic and computational analyses. 3meC analysis by LC/MS was carried out by R.A., C.E.R., S.L. and P.S. ESC CRISPR deletion and analysis was performed by A.G., J.K.R. and P.S. M.E. and A.J. carried out the in vitro DNMT3a analysis. T.B. and P.H.S. performed genome assembly. S.R., M.E.S., R.K.G. and A.J.B. were responsible for nematode culture. P.S., P.H. and A.J. analyzed the data and prepared the manuscript.

Competing interests

The authors declare no competing financial interests.

Corresponding authors

Correspondence to Petra Hajkova or Peter Sarkies.

Integrated supplementary information

  1. Supplementary Figure 1 Extended analysis of the sequence context of DNA methylation in nematodes

    a, Overall percentage methylation at CG and non-CG sites respectively for different nematodes. The chi-squared test P value is for a two-tailed test on a 2 by 2 table: meCG/non-meCG|meCX/non-meCX, where X is C, A or T. P = 1 × 10–100 is the minimum P value reported by the software used for calculation. See Supplemental Note 3 for the number of CGs in different contexts across the genome for all genomes analyzed. b,c, Enrichment of methylation for CG sites within different contexts for P. sambesii and T. spiralis respectively (nematodes with the most pronounced differences in CG[X] preferences; see Fig. 2). See Supplemental Note 3 for the number of CGs in different contexts across the genome for all genomes analyzed.

  2. Supplementary Figure 2

    Sample genome browser windows for P. sambesii and R. culicivorax showing lack of methylation on genes but high methylation on repeats

  3. Supplementary Figure 3

    Sample genome browser windows for T. spiralis and T. muris demonstrating methylation prominence at repeats rather than genes

  4. Supplementary Figure 4 Comparison of DNA methylation in genes and repeats

    ad, Methylation levels across genes for either genes with repeat homology or those without for the different nematode species with DNMT activity. P values shown are from two-tailed Wilcoxon unpaired tests. The box plot shows interquartile range, with a line at the median and whiskers extending to the most extreme point no more than 1.5 times the interquartile range from the box. The notch shows 95% confidence levels on the median (1.57 times the interquartile range/√number of samples). P values are reported to the nearest significant figure. Analysis is based on the number of genes reported in Supplemental Note 3.

  5. Supplementary Figure 5 Cytosine methylation and CG sequence content of genes in nematodes

    ad, Comparison of methylation as a fraction of total cytosine (rather than CG) and CG levels upstream and downstream of the transcriptional start site for the different nematode species with cytosine DNMT activity. e, Variation of CG methylation normalized to CG content with overall CG content for exons (diamonds) or introns (triangles) in different nematodes. The number of CG sites analyzed for all plots is in Supplemental Note 3.

  6. Supplementary Figure 6 Coevolution of DNA repair proteins with DNA methyltransferases

    Heat map showing the conservation of different proteins with a GO term for DNA repair that coevolve with cytosine DNA methylation. DNMT1 and DNMT3 are shown for comparison. In addition to alkylation repair, this set of genes includes components of base excision repair, which is also involved in repairing DNA alkylation damage, and double-strand break repair, which would be required in the event of replication fork stalling, for instance at 3meC lesions if not repaired by ALKB2.

  7. Supplementary Figure 7 ALKB2/3 coevolves with DNA methyltransferases across metazoa

    Heat map showing the conservation of ALKB and DNMT family members across metazoans. ALKB2/3 are the only ALKB family members that show significant coevolution with DNA methyltransferases.

  8. Supplementary Figure 8

    ALKB2/3 and DNMT coevolution in metazoa

  9. Supplementary Figure 9

    DNA methylation and ALKB2/3 coevolution in protists

  10. Supplementary Figure 10

    DNA methylation and ALKB2/3 coevolution in fungi

  11. Supplementary Figure 11 DNA methylation and ALKB2/3 coevolution in Ascomycota

    See also Fig. 9.

  12. Supplementary Figure 12 ALKB2/3 supports higher DNA methylation levels in arthropods

    DNA methylation levels across arthropods comparing 4 species with ALKB2/3 and 17 species without ALKB2/3. A coding sequence B across the genome. The only species that has ALKB2/3 yet shows extremely low levels of detectable methylation is T. castaneum. The box plot shows a line at the median, the interquartile range as the box and whiskers extending to the greatest point that is less than 1.5 times the interquartile range. For the full list of arthropod species analyzed, see Supplemental Table 9.

  13. Supplementary Figure 13 Establishment of a method to monitor 3meC by LC/MS

    a, Example of a standard curve and LCQ for 3meC. The standard curve was repeated each time measurements were taken (5), with similar results. b,c, 5meC and 3meC levels measured in mSssI methylated and unmethylated plasmid with or without exposure to MMS. Technical variation from two measurements is shown. Note low-level induction of 3meC by mSssI alone in c. The signal-to-noise ratio for 3meC detection was 10 as per the Methods. d,e, Screenshots of LC/MS traces for two further independent replicate in vitro reactions containing the catalytic domain of DNMT3a and an unmethylated plasmid, showing induction of 5meC and 3meC.

  14. Supplementary Figure 14 Analysis of alkylation damage caused by DNMTs

    a, Dot blot demonstrating the specificity of 3meC antibody, which recognizes a DNA template alkylated in vitro and not a 5meC-containing template synthesized by PCR, in contrast to the anti-5meC antibody, which only recognizes the 5meC template. The experiment was repeated twice with similar results. b, Dot blot showing 3meC levels in WT ES cells and cells carrying either DNMT3a and DNMT3b deletion (DKO) or DNMT3a, DNMT3b and DNMT1 deletion (TKO). The experiment was repeated four times with similar results. c, Western blot showing comparable protein levels for both WT and mutant (F646A) DNMT3a catalytic domain expressed and purified from E. coli. The experiment was performed twice with similar results.

  15. Supplementary Figure 15 Generation of clones carrying deletions in ALKB2 in mouse ES cells

    a, CRISPR strategy to generate disruptions in the first coding exon of mouse ALKBH2, the ortholog of ALKB2. b, Example of sequencing data showing out-of-frame deletion in both alleles, indicating successful disruption of the protein. c, Western blot confirming reduction of ALKB2 protein levels. Non-specific bands are indicated by an asterisk. Tubulin detection from the same membrane is shown as a control. The blot was performed three times for WT clones and two times for TKO clones. d, Sensitivity to 200 mM MMS. The y axis shows percentage surviving colonies relative to an untreated sample from the same line. The mean and range of two biological replicates are shown.

Supplementary information

  1. Supplementary Text and Figures

    Supplementary Figures 1–15, Supplementary Tables 1–9 and Supplementary Notes 1–3

  2. Life Sciences Reporting Summary

About this article

Publication history






Further reading