Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Molecular basis for DarT ADP-ribosylation of a DNA base


ADP-ribosyltransferases use NAD+ to catalyse substrate ADP-ribosylation1, and thereby regulate cellular pathways or contribute to toxin-mediated pathogenicity of bacteria2,3,4. Reversible ADP-ribosylation has traditionally been considered a protein-specific modification5, but recent in vitro studies have suggested nucleic acids as targets6,7,8,9. Here we present evidence that specific, reversible ADP-ribosylation of DNA on thymidine bases occurs in cellulo through the DarT–DarG toxin–antitoxin system, which is found in a variety of bacteria (including global pathogens such as Mycobacterium tuberculosis, enteropathogenic Escherichia coli and Pseudomonas aeruginosa)10. We report the structure of DarT, which identifies this protein as a diverged member of the PARP family. We provide a set of high-resolution structures of this enzyme in ligand-free and pre- and post-reaction states, which reveals a specialized mechanism of catalysis that includes a key active-site arginine that extends the canonical ADP-ribosyltransferase toolkit. Comparison with PARP–HPF1, a well-established DNA repair protein ADP-ribosylation complex, offers insights into how the DarT class of ADP-ribosyltransferases evolved into specific DNA-modifying enzymes. Together, our structural and mechanistic data provide details of this PARP family member and contribute to a fundamental understanding of the ADP-ribosylation of nucleic acids. We also show that thymine-linked ADP-ribose DNA adducts reversed by DarG antitoxin (functioning as a noncanonical DNA repair factor) are used not only for targeted DNA damage to induce toxicity, but also as a signalling strategy for cellular processes. Using M. tuberculosis as an exemplar, we show that DarT–DarG regulates growth by ADP-ribosylation of DNA at the origin of chromosome replication.

This is a preview of subscription content, access via your institution

Access options

Rent or buy this article

Prices vary by article type



Prices may be subject to local taxes which are calculated during checkout

Fig. 1: DarT is a PARP-like protein that catalyses ADP-ribosylation of DNA at the thymidine base nitrogen.
Fig. 2: DNA binding is highly coordinated by DarT for site-specific ADP-ribosylation of DNA.
Fig. 3: Mechanism of DNA ADP-ribosylation.
Fig. 4: ADP-ribosylation of gDNA in M. tuberculosis.

Similar content being viewed by others

Data availability

Crystallography atomic coordinates and structure factors are deposited in the PDB ( under the following accession codes: 7OMV, 7OMW, 7OMX, 7OMY, 7OMU, 7OMZ and 7ON0. RNA-sequencing sequence files are deposited at the NCBI Gene Expression Omnibus GEO under the accession code GSE174526. Tn-seq sequence files are deposited at the NCBI Sequence Read Archive, SRA accession number PRJNA532518 run SRR8886987. All data supporting the findings of this study are available within Article and any further information will be provided upon request to the corresponding authors.


  1. Barkauskaite, E., Jankevicius, G. & Ahel, I. Structures and mechanisms of enzymes employed in the synthesis and degradation of PARP-dependent protein ADP-ribosylation. Mol. Cell 58, 935–946 (2015).

    Article  CAS  Google Scholar 

  2. Gibson, B. A. & Kraus, W. L. New insights into the molecular and cellular functions of poly(ADP-ribose) and PARPs. Nat. Rev. Mol. Cell Biol. 13, 411–424 (2012).

    Article  CAS  Google Scholar 

  3. Cohen, M. S. & Chang, P. Insights into the biogenesis, function, and regulation of ADP-ribosylation. Nat. Chem. Biol. 14, 236–243 (2018).

    Article  CAS  Google Scholar 

  4. Perina, D. et al. Distribution of protein poly(ADP-ribosyl)ation systems across all domains of life. DNA Repair 23, 4–16 (2014).

    Article  CAS  Google Scholar 

  5. Kraus, W. L. PARPs and ADP-ribosylation: 50 years … and counting. Mol. Cell 58, 902–910 (2015).

    Article  CAS  Google Scholar 

  6. Talhaoui, I. et al. Poly(ADP-ribose) polymerases covalently modify strand break termini in DNA fragments in vitro. Nucleic Acids Res. 44, 9279–9295 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  7. Munnur, D. & Ahel, I. Reversible mono-ADP-ribosylation of DNA breaks. FEBS J. 284, 4002–4016 (2017).

    Article  CAS  Google Scholar 

  8. Munnur, D. et al. Reversible ADP-ribosylation of RNA. Nucleic Acids Res. 47, 5658–5669 (2019).

    Article  CAS  Google Scholar 

  9. Groslambert, J., Prokhorova, E. & Ahel, I. ADP-ribosylation of DNA and RNA. DNA Repair (Amst.) 105, 103144 (2021).

    Article  CAS  Google Scholar 

  10. Jankevicius, G., Ariza, A., Ahel, M. & Ahel, I. The toxin-antitoxin system DarTG catalyzes reversible ADP-ribosylation of DNA. Mol. Cell 64, 1109–1116 (2016).

    Article  CAS  Google Scholar 

  11. Aravind, L., Zhang, D., de Souza, R. F., Anand, S. & Iyer, L. M. in Endogenous ADP-ribosylation (ed. Koch-Nolte, F.) 3–32 (Springer International, 2014).

  12. Bell, C. E. & Eisenberg, D. Crystal structure of diphtheria toxin bound to nicotinamide adenine dinucleotide. Adv. Exp. Med. Biol. 419, 35–43 (1997).

    Article  CAS  Google Scholar 

  13. Steffen, J. D., Brody, J. R., Armen, R. S. & Pascal, J. M. Structural implications for selective targeting of PARPs. Front. Oncol. 3, 301 (2013).

    Article  Google Scholar 

  14. Vyas, S. et al. Family-wide analysis of poly(ADP-ribose) polymerase activity. Nat. Commun. 5, 4426 (2014).

    Article  ADS  CAS  Google Scholar 

  15. Hottiger, M. O., Hassa, P. O., Lüscher, B., Schüler, H. & Koch-Nolte, F. Toward a unified nomenclature for mammalian ADP-ribosyltransferases. Trends Biochem. Sci. 35, 208–219 (2010).

    Article  CAS  Google Scholar 

  16. Belousova, E. A., Ishchenko, А. A. & Lavrik, O. I. DNA is a new target of Parp3. Sci. Rep. 8, 4176 (2018).

    Article  ADS  CAS  Google Scholar 

  17. Zarkovic, G. et al. Characterization of DNA ADP-ribosyltransferase activities of PARP2 and PARP3: new insights into DNA ADP-ribosylation. Nucleic Acids Res. 46, 2417–2431 (2018).

    Article  CAS  Google Scholar 

  18. Ueda, K. & Hayaishi, O. ADP-ribosylation. Annu. Rev. Biochem. 54, 73–100 (1985).

    Article  CAS  Google Scholar 

  19. Lawarée, E. et al. DNA ADP-ribosylation stalls replication and is reversed by RecF-mediated homologous recombination and nucleotide excision repair. Cell Rep. 30, 1373–1384 (2020).

    Article  Google Scholar 

  20. Mendum, T. A. et al. Transposon libraries identify novel Mycobacterium bovis BCG genes involved in the dynamic interactions required for BCG to persist during in vivo passage in cattle. BMC Genomics 20, 431 (2019).

    Article  Google Scholar 

  21. Ahel, I. et al. The neurodegenerative disease protein aprataxin resolves abortive DNA ligation intermediates. Nature 443, 713–716 (2006).

    Article  ADS  CAS  Google Scholar 

  22. Tumbale, P. et al. Structure of an aprataxin–DNA complex with insights into AOA1 neurodegenerative disease. Nat. Struct. Mol. Biol. 18, 1189–1195 (2011).

    Article  CAS  Google Scholar 

  23. Harms, A., Brodersen, D. E., Mitarai, N. & Gerdes, K. Toxins, targets, and triggers: an overview of toxin-antitoxin biology. Mol. Cell 70, 768–784 (2018).

    Article  CAS  Google Scholar 

  24. Gamulin, V., Cetkovic, H. & Ahel, I. Identification of a promoter motif regulating the major DNA damage response mechanism of Mycobacterium tuberculosis. FEMS Microbiol. Lett. 238, 57–63 (2004).

    CAS  PubMed  Google Scholar 

  25. Suskiewicz, M. J. et al. HPF1 completes the PARP active site for DNA damage-induced ADP-ribosylation. Nature 579, 598–602 (2020).

    Article  ADS  CAS  Google Scholar 

  26. Singh, A. K. et al. Investigating essential gene function in Mycobacterium tuberculosis using an efficient CRISPR interference system. Nucleic Acids Res. 44, e143 (2016).

    Article  Google Scholar 

  27. Larson, M. H. et al. CRISPR interference (CRISPRi) for sequence-specific control of gene expression. Nat. Protocols 8, 2180–2196 (2013).

    Article  CAS  Google Scholar 

  28. St-Pierre, F. et al. One-step cloning and chromosomal integration of DNA. ACS Synth. Biol. 2, 537–541 (2013).

    Article  CAS  Google Scholar 

  29. Thu, K. L. et al. Methylated DNA immunoprecipitation. J. Vis. Exp. 935, 935 (2009).

    Google Scholar 

  30. Winter, G. xia2: an expert system for macromolecular crystallography data reduction. J. Appl. Crystallogr. 43, 186–190 (2010).

    Article  CAS  Google Scholar 

  31. Storoni, L. C., McCoy, A. J. & Read, R. J. Likelihood-enhanced fast rotation functions. Acta Crystallogr. D 60, 432–438 (2004).

    Article  Google Scholar 

  32. Cowtan, K. Recent developments in classical density modification. Acta Crystallogr. D 66, 470–478 (2010).

    Article  CAS  Google Scholar 

  33. Cowtan, K. The Buccaneer software for automated model building. 1. Tracing protein chains. Acta Crystallogr. D 62, 1002–1011 (2006).

    Article  Google Scholar 

  34. Emsley, P. & Cowtan, K. Coot: model-building tools for molecular graphics. Acta Crystallogr. D 60, 2126–2132 (2004).

    Article  Google Scholar 

  35. Murshudov, G. N., Vagin, A. A. & Dodson, E. J. Refinement of macromolecular structures by the maximum-likelihood method. Acta Crystallogr. D 53, 240–255 (1997).

    Article  CAS  Google Scholar 

  36. Chen, V. B. et al. MolProbity: all-atom structure validation for macromolecular crystallography. Acta Crystallogr. D 66, 12–21 (2010).

    Article  CAS  Google Scholar 

  37. Waterhouse, A. M., Procter, J. B., Martin, D. M. A., Clamp, M. & Barton, G. J. Jalview Version 2—a multiple sequence alignment editor and analysis workbench. Bioinformatics 25, 1189–1191 (2009).

    Article  CAS  Google Scholar 

  38. Katoh, K., Rozewicki, J. & Yamada, K. D. MAFFT online service: multiple sequence alignment, interactive sequence choice and visualization. Brief. Bioinform. 20, 1160–1166 (2019).

    Article  CAS  Google Scholar 

  39. Saitou, N. & Nei, M. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406–425 (1987).

    CAS  PubMed  Google Scholar 

  40. Ashkenazy, H. et al. ConSurf 2016: an improved methodology to estimate and visualize evolutionary conservation in macromolecules. Nucleic Acids Res. 44, W344–W350 (2016).

    Article  CAS  Google Scholar 

  41. Tsurumura, T. et al. Arginine ADP-ribosylation mechanism based on structural snapshots of iota-toxin and actin complex. Proc. Natl Acad. Sci. USA 110, 4267–4272 (2013).

    Article  ADS  CAS  Google Scholar 

Download references


We thank C. Tang for guidance and discussion of this study and sharing reagents; J. Rack and M. Suskiewicz for discussions and critically revising the manuscript; M. Suskiewicz and E. Lowe for help with X-ray diffraction data collection and processing; S. Meyer for help with biochemical experiments; J. Elkins for discussions; J. Holder and T. Agnew for proofreading of the manuscript; T. Mendum, H. Wu and A. Smith for help with transposon mutagenesis; S. Hingley-Wilson for discussions on toxin–antitoxin systems; and Diamond Light Source for access to and assistance at beamlines I03, I04, I04-1, and I24 throughout the project (proposal numbers mx18069). Work in the laboratory of I.A. is supported by the Wellcome Trust (101794 and 210634), Biotechnology and Biological Sciences Research Council (BB/R007195/1) and Cancer Research United Kingdom (C35050/A22284). The G.R.S. and S.L.K. laboratories were supported by Biotechnology and Biological Sciences Research Council grants BB/R006393/1 and BB/N004590/1, respectively.

Author information

Authors and Affiliations



I.A. and G.R.S. conceived the project and conceptualized experiments with input from M.S., R.E.B. and G.J. M.S. conducted biochemical and crystallographic studies, including structure and data analysis and interpretation, with assistance of other authors; A.A. solved T. africanus DarTG structure and refined structural data; C.T.-C. and G.J. established method for detection of ADP-ribosylated DNA and supported strain construction; T.D.W.C. conducted NMR experiments and analysis; R.E.B. and G.R.S. performed mycobacteria experiments with assistance from S.L.K. and S.G. for DarG knockdown. M.S., I.A., G.R.S. and R.E.B. wrote the manuscript with support of all other authors.

Corresponding authors

Correspondence to Graham R. Stewart or Ivan Ahel.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature thanks the anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.

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

Extended data figures and tables

Extended Data Fig. 1 DarT structure reveals a PARP-like ADP-ribosyltransferase.

Related to Fig. 1. a, Crystal structure of T. africanus DarTG(E152A) fusion protein served as model for solving DarT of Thermus sp. 2.9 by molecular replacement. The fused DarG macrodomain is coloured in light orange with the bound ADP-ribose molecule shown as an atom-coloured stick model in black. b, c, Comparison of DarT with eukaryotic ARTD (that is, PARP) and bacterial ARTD fold. b, Secondary structure analysis shows the close similarity of DarT to ARTD family members, and PARPs in particular, with its fold-stabilizing central 6-stranded β-sheet core and the ARTD-conserved helices between strands β1 and β2 (β1-2) and β2 and β3 (β2-3). The crystal structure of Thermus sp. 2.9 DarT(E160A) in ligand-free state (apo) was overlaid with diphtheria toxin (PDB code 1TOX), PARP1 (PDB code 6BHV) and PARP13 (PDB code 2X5Y). For clarity, only central secondary structure elements showing the similarity between the folds are depicted from diphtheria toxin, PARP1 and PARP13. c, The N-terminal extension of the β-sheet core found in PARPs (that is, a strand–helix–strand arrangement next to β6) is spatially replaced in DarT with a shorter C-terminal helix–strand extension. The crystal structure of Thermus sp. 2.9 DarT(E160A) in ADPr–DNA-bound state was overlaid with the crystal structure of PARP1 (PDB code 6BHV) (left) and PARP13 (PDB code 2X5Y) (right). For clarity, only the secondary structure elements showing this difference between the folds are depicted from PARP1 and PARP13. d, Thermus sp. 2.9 DarT(E160A) in complex with NAD+ and carba-NAD+. Overlay of the crystal structures with Thermus sp. 2.9 DarT(E160A) in ligand-free state (apo) is shown on the right. Unresolved regions of the NAD+-binding loop-helical element including the ART donor loop (purple) in the DarT–NAD+ and DarT–carba-NAD+ structures are marked with asterisks.

Extended Data Fig. 2 Structural details of ADPr–DNA bound to DarT.

Related to Fig. 1c, d. a, Crystal structure of Thermus sp. 2.9 DarT(E160A) in complex with ADPr–DNA. Co-crystallization with ADP-ribosylated DNA 5-mer after in vitro modification by T. aquaticus DarT and purification. The substrate-binding ARTT loop is highlighted in green, the NAD+-binding loop–helix element in purple (set, for clarity, in higher transparency). The catalytic glutamate E160, which is conserved in ARTs, is modelled as red sticks. b, Overlay of the ADP-ribosylated DNA products of the ADPr–DNA- and NAM-bound and the ADPr–DNA-bound DarT structures shows their perfect overlap, apart from a slight tilting of the bonds that connect the NAM-ribose with the β-phosphate. c, d, The ADPr–DNA ligands are highly resolved in the Thermus sp. 2.9 DarT(E160A) co-crystal structures, revealing the ADP-ribose linkage to the thymidine base nitrogen N3 in α configuration. The 2Fo − Fc electron density maps contoured at 1.0σ around the ligands are shown in blue. ADPr, linked ADP-ribose; NAM, nicotinamide; ssDNA, single-stranded DNA. c, The ADPr–DNA ligand in the ADPr–DNA co-crystal structure of 1.46 Å resolution. d, The ADPr–DNA ligand in the ADPr–DNA- and NAM-bound co-crystal structure at 1.66 Å resolution. The NAM ligand left in the protein after ADP-ribosylation of the DNA is also clearly resolved. e, ADP-ribosylation activity of the T. aquaticus DarT(E160A) mutant can also be observed in in vitro assays at low DNA (50 nM) and high protein concentrations under long incubation times. Modification of the Cy3-labelled oligonucleotide (DarT-ADPr-27mer-Cy3) was visualized after separation of the reaction products on denaturing polyacrylamide gel. Representative of two independent experiments.

Extended Data Fig. 3 Structural features of DarT for ssDNA binding and catalysis.

Related to Figs. 13. a, The ADP-ribosylating turn–turn (ARTT) loop of Thermus sp. 2.9 DarT(E160A) in the ADPr–DNA- and NAM-bound structure is shown in green with its stabilized DNA substrate in magenta. Several loops form together with the few short α-helices a stable scaffold, which is held in position by a network of over 100 interactions between main chains, side chains and water molecules. Left, cartoon representation; middle, the atom-coloured stick model of the ARTT loop. Interactions are indicated with grey dashes and water molecules as red spheres. Right, table comparing the ARTT loop length of DarT with other human and bacterial ARTDs15. b, Thermus sp. 2.9 DarT preferentially modifies a TNTC motif in ssDNA, which was verified by testing permutations of the motif. In vitro ADP-ribosylation activity of Thermus sp. 2.9 DarT was assessed by visualizing the modification of the oligonucleotides under UV light after separation and ethidium bromide staining of the reaction products on denaturing polyacrylamide gel. Representative of three independent experiments. c, Close views on the nucleotide recognition of DarT, rationalizing its preferred modification of DNA over RNA. Cartoon–stick models of the Thermus sp. 2.9 DarT(E160A) structure in the ADPr–DNA- and NAM-bound state are shown. Left, middle, additional 2′ hydroxyl groups as in RNA strands may lead to clashes with parts of the proteins (that is, W147 (1st nucleotide) and the α-helix between β2 and β3 (2nd nucleotide)). Right, as shown in previous studies10, the methyl group on the modified thymine (circle) increases thymidine base modification, probably by locking the base in optimal conformation for the ADP-ribosylation reaction. Interactions are indicated with grey dashes and water molecules as red spheres. d, Modelling of possible rotamers of glutamate E160 into the Thermus sp. 2.9 DarT apo structure. Several conformations of the glutamate would allow a proton transfer from arginine R51 (green) to glutamate E160 (red). Possible interactions are shown with dashes in magenta. e, Arginine R51 flexibility observed among the Thermus sp. 2.9 DarT apo, substrate- and product-bound states. The NAD+- and DNA-bound and as ADP-ribose unlinked state (fourth imagine from left) is modelled by superimposing the NAD+ molecule with the NAD+ co-crystal structure onto the carba-NAD+- and DNA-bound structure. R51 and ligands are shown as atom-coloured stick models, with R51 in green, NAD+ in cyan, carba-NAD+ in brown and higher transparency and DNA (thymine only) in magenta. Interactions are indicated with grey dashes.

Extended Data Fig. 4 DarT sequence alignments.

Related to Figs. 2c, d, 3. a, Sequence alignment of Thermus sp. 2.9 DarT with DarT of T. aquaticus. Numbers on top of the alignments refer to Thermus sp. 2.9 DarT. Table provides a residue identifier comparison for functional relevant residues. b, Multiple sequence alignment of DarT sequences, representing five main phylogenetically diverging branches. Numbers on top of the residues refer to Thermus sp. 2.9 DarT. Active site residues are highlighted in green, DNA-binding residues in magenta, with functionally similar residues as the reference in lower opacity. Shared sequence identities compared to Thermus sp. 2.9 DarT: group 1: 60%, group 2: 40–45%, group 3 and 4: 31–38%, group 5: 20–27%.

Extended Data Fig. 5 NAD+ coordination in the active site of DarT.

Related to Fig. 3, Supplementary Discussion. a, Comparison of the NAD+-binding sites in the carba-NAD+-bound structure and the NAD+-bound structure of Thermus sp. 2.9 DarT(E160A). Top, Overlay of the carba-NAD+- and DNA-bound structure (grey) with the carba-NAD+-bound structure (brown), of which just the ligand and the side chains are shown as an atom-coloured stick model. The carba-NAD+ ligands of both structures perfectly overlay and DarT-interacting side chains show same positioning. DNA binding does not induce conformational changes upon the NAD+ ligand. Bottom, overlay of the carba-NAD+- and DNA-bound structure (grey) with the NAD+-bound structure (cyan) shows slight differences in the ligand and the side chains positioning around the pyrophosphate-ribose moiety of the NAD+ molecule, which needs to be considered for analysis of NAD+ polarization. b, Molecular structures of NAD+ and carba-NAD+. c, Cartoon–stick model showing the coordination of the NAM side (left) and the adenine side (right) of the carba-NAD+ ligand in the Thermus sp. 2.9 DarT(E160A) carba-NAD+- and DNA-bound structure with side- and main-chain interactions (dashed lines), including water (red spheres) contacts. d, Integrated thermogram obtained by isothermal titration calorimetry giving NAD+-binding parameters for Thermus sp. 2.9 DarT(E160Q). A representative result from three independent experiments is shown, with the number of binding sites (n) and the dissociation constant (KD) calculated from the repeats with mean ± s.d. See Supplementary Fig. 3 for raw titration curves and additional thermodynamic parameters. e, Autoradiography of thin layer chromatography plate analysing the reaction products after incubation of T. aquaticus DarT and DarT(E160A) with NAD+ and DNA. NADase from pig brain was used as control for monitoring NADase activity. Representative of three independent experiments.

Extended Data Fig. 6 Visualization of the ADP-ribosylation activity of DarT in cells.

Related to Fig. 4. a, Validation of the antibody identified for detection of ADP-ribosylated DNA. ADP-ribosylation of the oligonucleotide by T. aquaticus DarT was verified by analysis of the reaction product on denaturing polyacrylamide gel (top) and visualized by immunoblotting using the poly/mono-ADP ribose antibody, E6F6A (Cell Signaling Technology) (bottom). Immunodetection of ssDNA (using the DSHB autoanti-ssDNA antibody) served as loading control (middle). Representative result of four independent experiments with three individually purified T. aquaticus DarT-ADP-ribosylated oligonucleotides. b, Dot blot showing DNA ADP-ribosylation activity by T. aquaticus and EPEC wild-type DarT and mutants on gDNA, its physiological target, (row 1 and 2 from top), which consequently induces DNA damage (RecA marker) in cells. EPEC DarT(G49D) is a characterized DarT mutant that retains ssDNA ADP-ribosylation activity, albeit to a lesser extent than the wild-type protein, and EPEC DarT(E170) is its respective catalytically inactive mutant19. See c for the structural basis of this lower activity of EPEC DarT(G49D). c, G49D mutation in EPEC DarT reduces ssDNA ADP-ribosylation activity. Overlay of a homology model of EPEC DarT with the structure of Thermus sp. 2.9 DarT(E160A) in complex with ADPr–DNA indicates that the EPEC DarT mutation G49D reduces DarT ssDNA activity owing to an aspartate side chain pointing into the NAD+-binding site towards the second phosphate group. This may sterically, but also owing to its negative charge, impair NAD+ binding, resulting in a less efficient ADP-ribosylation reaction. d, DNA ADP-ribosylation by T. aquaticus DarT (dot blot, row 1 and 2 from top) and induction of DNA damage (RecA marker) is suppressed by T. aquaticus DarG with its macrodomain (MD), including by DarG macrodomains from non-cognate species (EPEC and M. tuberculosis). e, Dot blot showing ADP-ribose removal from T. aquaticus DarT-ADP-ribosylated gDNA by T. aquaticus DarG antitoxin with its macrodomain, and macrodomains from non-cognate species (EPEC and M. tuberculosis) in contrast to the human hydrolases MacroD1, PARG and ARH3. In b, d, e, cell lysates were prepared and gDNA was purified from samples before (+ glucose) and after (+ arabinose/IPTG) induction of protein expression and subjected to immunodetection. EV, empty vector. The N22A/K80A double mutation in T. aquaticus DarG results in loss of catalytic activity of the macrodomain. For gel source data, see Supplementary Fig. 1. Results are representative for three biologically independent experiments.

Extended Data Fig. 7 Characterization of DarT gDNA ADP-ribosylation in M. tuberculosis.

Related to Fig. 4. a, Unregulated DarT activity (darG silencing) and induction of DNA damage (MMC treatment) led to a marked DNA damage response and induces expression of dnaB-darT. Gene transcription was compared by qPCR with reverse transcription of M. bovis BCG darG sgRNA un-induced, ATC-induced and MMC-treated samples. Data are mean ± s.d. of three biologically independent replicates. b, Knockdown of darG expression in M. tuberculosis induces expression of RecA. Mycobacterium tuberculosis was treated with 200 ng ml−1 ATC to induce dCas9 and darG sgRNA or non-targeting control sgRNA for 48 h or with MMC for 24 h. Cell-free bacterial lysates were probed by western blotting with an anti-RecA antiserum or anti-Hsp70 (DnaK) antibody as loading control. Representative of two biologically independent experiments. c, darT and darG are transcriptionally linked to dnaB. PCR products were generated with the indicated set of primers (Supplementary Table 2) and visualized by gel electrophoresis. The presence of PCR products across the dnaB-darT and darT-darG junctions demonstrates the transcriptional linkage of dnaB, darT and darG as a polycistronic mRNA. Representative of three independent experiments. d, Mycobacterium tuberculosis DarT preferentially modifies a TTTW motif in ssDNA. Screening of 40 ssDNA oligonucleotide sequences with potential four-base motifs for ADP-ribosylation by DarT (data not shown) identified TTTW as targeted sequence, which was verified by testing permutations of the TTTT motif. In vitro ADP-ribosylation activity of M. tuberculosis DarT was assessed by visualizing the modification of the oligonucleotides under UV light after separation and ethidium bromide staining of the reaction products on denaturing polyacrylamide gel. Representative of three independent experiments. e, Mycobacterium tuberculosis DarT ADP-ribosylates the OriC in vitro with preference for the lower strand at the TTTW motifs. ADP-ribosylation activity was assessed by visualizing the modification of the oligonucleotides under UV light after separation and ethidium bromide staining of the reaction products on denaturing polyacrylamide gel. Representative of three independent experiments. For gel source data, see Supplementary Fig. 1.

Extended Data Fig. 8 DarT is a PARP-like protein that evolved novel features that allow its specialized function as a DNA ADP-ribosyltransferase.

Related to ‘Discussion’ in the Article. a, Schematics of the interactions between the NAD+ substrate and the residues of the class-defining (H-Y-E) motif in ARTD members, including PARPs, compared to DarT. Conserved motif residues (purple) and additional active site residues (green) essential for catalysis with their relative position to the NAD+ substrate are compared. b, ARTs seem to share the spatial position and orientation of mechanistically relevant residues. Overlay of crystal structures of Thermus sp. 2.9 DarT(E160A) in ADPr–DNA- and NAM-bound state with Clostridium perfringens iota-toxin (Ia)–actin complex (left) (PDB code 4H0T) and PARP2 in the PARP2–HPF1 complex (right) (PDB code 6TX3). H119 in DarT takes spatially the same position as Y375 in the iota-toxin, which was suggested to have a role in target protein (that is, actin) recognition41. Both Y375 and H119 are accommodated in the ARTT loops, which do not show any similarity in either residue length or structural make up. The approximate position of DarT H119 is occupied by E284 of HPF1 in the PARP2–HPF1 complex, in which HPF1 sits on the ARTT loop of PARP2. This leads to the formation of a composite active site with the catalytic glutamate residues E284 and E545 for catalysing serine ADP-ribosylation25. Enlarged views of the active sites are below the respective cartoon models. For clarity in the enlarged views, only the ARTT loop from iota-toxin and PARP2 and only a fragment of the respective binding partner (that is, actin and HPF1) are shown as cartoon model. The substrate-coordinating and catalytic residues as well as the ADP-ribose products and complex-bound ligands are shown as stick models. Root mean square deviation (r.m.s.d.) of DarT–iota-toxin overlay, 2.71 Å; r.m.s.d. of DarT–PARP2 overlay, 2.58 Å.

Extended Data Table 1 Data collection and refinement statistics for crystal structures of T. africanus DarTG and Thermus sp. 2.9 DarT described in this Article
Extended Data Table 2 Data collection and refinement statistics for DNA co-crystal structures of Thermus sp. 2.9 DarT described in this Article

Supplementary information

Supplementary Information

This file contains Supplementary Figs 1-4, Supplementary Tables 1-3 and Supplementary References.

Reporting Summary

Supplementary Data

RNA-Seq Differential Expression Analysis.

Peer Review File

Rights and permissions

Reprints and Permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Schuller, M., Butler, R.E., Ariza, A. et al. Molecular basis for DarT ADP-ribosylation of a DNA base. Nature 596, 597–602 (2021).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

This article is cited by


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing