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.

Time-resolved structural analysis of an RNA-cleaving DNA catalyst


The 10–23 DNAzyme is one of the most prominent catalytically active DNA sequences1,2. Its ability to cleave a wide range of RNA targets with high selectivity entails a substantial therapeutic and biotechnological potential2. However, the high expectations have not yet been met, a fact that coincides with the lack of high-resolution and time-resolved information about its mode of action3. Here we provide high-resolution NMR characterization of all apparent states of the prototypic 10–23 DNAzyme and present a comprehensive survey of the kinetics and dynamics of its catalytic function. The determined structure and identified metal-ion-binding sites of the precatalytic DNAzyme–RNA complex reveal that the basis of the DNA-mediated catalysis is an interplay among three factors: an unexpected, yet exciting molecular architecture; distinct conformational plasticity; and dynamic modulation by metal ions. We further identify previously hidden rate-limiting transient intermediate states in the DNA-mediated catalytic process via real-time NMR measurements. Using a rationally selected single-atom replacement, we could considerably enhance the performance of the DNAzyme, demonstrating that the acquired knowledge of the molecular structure, its plasticity and the occurrence of long-lived intermediate states constitutes a valuable starting point for the rational design of next-generation DNAzymes.

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

Relevant articles

Open Access articles citing this article.

Access options

Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Fig. 1: The precatalytic Dz–RNA complex adopts an unexpected but effective structure that is in dynamic exchange with divalent metal ions.
Fig. 2: Metal-ion interactions determine the activity of Dz by condensing the structure and bringing the catalytic loop into an active conformation.
Fig. 3: The catalytic reaction involves multiple rate-limiting transient intermediate states that offer suitable targets to increase the activity of Dz.

Data availability

NMR resonance assignments of Dz5C–RNA2ʹF are available via the BMRB with the accession code: 34654. Atomic coordinates of the Dz5C–RNA2ʹF ensemble (cluster I) are deposited in the PDB under the accession code: 7PDU. Coordinates of additional cluster and scripts for data analysis are available from the corresponding author upon reasonable request.

Code availability

The codes used for structure calculation, MD simulations and simulation of real-time NMR data are available from the corresponding author upon reasonable request.


  1. Santoro, S. W. & Joyce, G. F. A general purpose RNA-cleaving DNA enzyme. Proc. Natl Acad. Sci. USA 94, 4262–4266 (1997).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  2. Silverman, S. K. Catalytic DNA: scope, applications, and biochemistry of deoxyribozymes. Trends Biochem. Sci 41, 595–609 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  3. Rosenbach, H. et al. Molecular features and metal ions that influence 10-23 DNAzyme activity. Molecules 25, 3100 (2020).

    CAS  PubMed Central  Google Scholar 

  4. Cuenoud, B. & Szostak, J. W. A DNA metalloenzyme with DNA ligase activity. Nature 375, 611–614 (1995).

    ADS  CAS  PubMed  Google Scholar 

  5. Pradeepkumar, P. I., Höbartner, C., Baum, D. A. & Silverman, S. K. DNA-catalyzed formation of nucleopeptide linkages. Angew. Chemie Int. Ed. 47, 1753–1757 (2008).

    CAS  Google Scholar 

  6. Shen, Z. et al. A catalytic DNA activated by a specific strain of bacterial pathogen. Angew. Chemie Int. Ed. 55, 2431–2434 (2016).

    CAS  Google Scholar 

  7. Torabi, S.-F. et al. In vitro selection of a sodium-specific DNAzyme and its application in intracellular sensing. Proc. Natl Acad. Sci. USA 112, 5903–5908 (2015).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  8. Ponce-Salvatierra, A., Wawrzyniak-Turek, K., Steuerwald, U., Höbartner, C. & Pena, V. Crystal structure of a DNA catalyst. Nature 529, 231–234 (2016).

    ADS  CAS  PubMed  Google Scholar 

  9. Liu, H. et al. Crystal structure of an RNA-cleaving DNAzyme. Nat. Commun. 8, 2006 (2017).

    ADS  PubMed  PubMed Central  Google Scholar 

  10. Victor, J., Steger, G. & Riesner, D. Inability of DNAzymes to cleave RNA in vivo is due to limited Mg2+ concentration in cells. Eur. Biophys. J. 47, 333–343 (2018).

    CAS  PubMed  Google Scholar 

  11. Young, D. D., Lively, M. O. & Deiters, A. Activation and deactivation of DNAzyme and antisense function with light for the photochemical regulation of gene expression in mammalian cells. J. Am. Chem. Soc. 132, 6183–6193 (2010).

    CAS  PubMed  PubMed Central  Google Scholar 

  12. Wang, Y., Nguyen, K., Spitale, R. C. & Chaput, J. C. A biologically stable DNAzyme that efficiently silences gene expression in cells. Nat. Chem. 13, 319–326 (2021).

    CAS  PubMed  Google Scholar 

  13. Liu, C. et al. Self‐assembly of copper–DNAzyme nanohybrids for dual‐catalytic tumor therapy. Angew. Chemie Int. Ed. 60, 14324–14328 (2021).

    CAS  Google Scholar 

  14. Wang, Y., Liu, E., Lam, C. H. & Perrin, D. M. A densely modified M2+-independent DNAzyme that cleaves RNA efficiently with multiple catalytic turnover. Chem. Sci. 9, 1813–1821 (2018).

    CAS  PubMed  PubMed Central  Google Scholar 

  15. Nowakowski, J., Shim, P. J., Prasad, G. S., Stout, C. D. & Joyce, G. F. Crystal structure of an 82-nucleotide RNA-DNA complex formed by the 10-23 DNA enzyme. Nat. Struct. Biol. 6, 151–156 (1999).

    CAS  PubMed  Google Scholar 

  16. Zaborowska, Z., Fürste, J. P., Erdmann, V. A. & Kurreck, J. Sequence requirements in the catalytic core of the “10–23” DNA enzyme. J. Biol. Chem. 277, 40617–40622 (2002).

    CAS  PubMed  Google Scholar 

  17. Vögeli, B., Olsson, S., Güntert, P. & Riek, R. The exact NOE as an alternative in ensemble structure determination. Biophys. J. 110, 113–126 (2016).

    ADS  PubMed  PubMed Central  Google Scholar 

  18. Elstner, M., Hobza, P., Frauenheim, T., Suhai, S. & Kaxiras, E. Hydrogen bonding and stacking interactions of nucleic acid base pairs: a density-functional-theory based treatment. J. Chem. Phys. 114, 5149–5155 (2001).

    ADS  CAS  Google Scholar 

  19. Kisseleva, N., Khvorova, A., Westhof, E. & Schiemann, O. Binding of manganese(II) to a tertiary stabilized hammerhead ribozyme as studied by electron paramagnetic resonance spectroscopy. RNA 11, 1–6 (2005).

    CAS  PubMed  PubMed Central  Google Scholar 

  20. Rosenbach, H. et al. Influence of monovalent metal ions on metal binding and catalytic activity of the 10–23 DNAzyme. Biol. Chem. 402, 99–111 (2020).

    PubMed  Google Scholar 

  21. Balbach, J. et al. Protein folding monitored at individual residues during a two-dimensional NMR experiment. Science 274, 1161–1163 (1996).

    ADS  CAS  PubMed  Google Scholar 

  22. Etzkorn, M., Böckmann, A. & Baldus, M. Kinetic analysis of protein aggregation monitored by real-time 2D solid-state NMR spectroscopy. J. Biomol. NMR 49, 121–129 (2011).

    CAS  PubMed  PubMed Central  Google Scholar 

  23. Nawrot, B. et al. Mapping of the functional phosphate groups in the catalytic core of deoxyribozyme 10-23. FEBS J. 274, 1062–1072 (2007).

    CAS  PubMed  Google Scholar 

  24. Horton, T. E., Clardy, D. R. & DeRose, V. J. Electron paramagnetic resonance spectroscopic measurement of Mn2+ binding affinities to the hammerhead ribozyme and correlation with cleavage activity. Biochemistry 37, 18094–18101 (1998).

    CAS  PubMed  Google Scholar 

  25. Hunsicker-Wang, L., Vogt, M. & Derose, V. J. EPR methods to study specific metal-ion binding sites in RNA. Methods Enzymol. 468, 335–367 (2009).

    CAS  PubMed  Google Scholar 

  26. Rinnenthal, J., Klinkert, B., Narberhaus, F. & Schwalbe, H. Modulation of the stability of the Salmonella fourU-type RNA thermometer. Nucleic Acids Res. 39, 8258–8270 (2011).

    CAS  PubMed  PubMed Central  Google Scholar 

  27. Delaglio, F. et al. NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J. Biomol. NMR 6, 277–293 (1995).

    CAS  PubMed  Google Scholar 

  28. Keller, R. L. J. Computer Aided Resonance Assignment Tutorial (Cantina, 2004).

  29. Lee, W., Tonelli, M. & Markley, J. L. NMRFAM-SPARKY: enhanced software for biomolecular NMR spectroscopy. Bioinformatics 31, 1325–1327 (2015).

    PubMed  Google Scholar 

  30. Schubert, S. RNA cleaving ‘10-23’ DNAzymes with enhanced stability and activity. Nucleic Acids Res. 31, 5982–5992 (2003).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  31. Thakur, C. S., Luo, Y., Chen, B., Eldho, N. V. & Dayie, T. K. Biomass production of site selective 13C/15N nucleotides using wild type and a transketolase E. coli mutant for labeling RNA for high resolution NMR. J. Biomol. NMR 52, 103–114 (2012).

    CAS  PubMed  Google Scholar 

  32. Orts, J., Vögeli, B. & Riek, R. relaxation matrix analysis of spin diffusion for the NMR structure calculation with eNOEs. J. Chem. Theory Comput. 8, 3483–3492 (2012).

    CAS  PubMed  Google Scholar 

  33. Schwieters, C. D., Kuszewski, J. J., Tjandra, N. & Marius Clore, G. The Xplor-NIH NMR molecular structure determination package. J. Magn. Reson. 160, 65–73 (2003).

    ADS  CAS  PubMed  Google Scholar 

  34. Valafar, H. & Prestegard, J. H. REDCAT: a residual dipolar coupling analysis tool. J. Magn. Reson. 167, 228–241 (2004).

    ADS  CAS  PubMed  Google Scholar 

  35. Huang, K. et al. Solution structure of the MEF2A–DNA complex: structural basis for the modulation of DNA bending and specificity by MADS-box transcription factors. EMBO J. 19, 2615–2628 (2000).

    CAS  PubMed  PubMed Central  Google Scholar 

  36. Pimenta, J. et al. NMR solution structure and SRP54M predicted interaction of the N-terminal sequence (1-30) of the ovine Doppel protein. Peptides 49, 32–40 (2013).

    CAS  PubMed  Google Scholar 

  37. Case, D. A. et al. Amber 2020 (Univ. California, 2020).

  38. Wang, J., Cieplak, P. & Kollman, P. A. How well does a restrained electrostatic potential (RESP) model perform in calculating conformational energies of organic and biological molecules? J. Comput. Chem. 21, 1049–1074 (2000).

    CAS  Google Scholar 

  39. Bas, D. C., Rogers, D. M. & Jensen, J. H. Very fast prediction and rationalization of pKa values for protein-ligand complexes. Proteins 73, 765–783 (2008).

    CAS  PubMed  Google Scholar 

  40. Jorgensen, W. L., Chandrasekhar, J., Madura, J. D., Impey, R. W. & Klein, M. L. Comparison of simple potential functions for simulating liquid water. J. Chem. Phys. 79, 926–935 (1983).

    ADS  CAS  Google Scholar 

  41. Joung, I. S. & Cheatham, T. E. Determination of alkali and halide monovalent ion parameters for use in explicitly solvated biomolecular simulations. J. Phys. Chem. B 112, 9020–9041 (2008).

    CAS  PubMed  PubMed Central  Google Scholar 

  42. Schott-Verdugo, S. & Gohlke, H. PACKMOL-Memgen: a simple-to-use, generalized workflow for membrane-protein–lipid-bilayer system building. J. Chem. Inf. Model. 59, 2522–2528 (2019).

    CAS  PubMed  Google Scholar 

  43. Li, P., Roberts, B. P., Chakravorty, D. K. & Merz, K. M. Rational design of particle mesh Ewald compatible Lennard-Jones parameters for +2 metal cations in explicit solvent. J. Chem. Theory Comput. 9, 2733–2748 (2013).

    CAS  PubMed  PubMed Central  Google Scholar 

  44. Darden, T., York, D. & Pedersen, L. Particle mesh Ewald: an N log(N) method for Ewald sums in large systems. J. Chem. Phys. 98, 10089–10092 (1993).

    ADS  CAS  Google Scholar 

  45. Ryckaert, J.-P., Ciccotti, G. & Berendsen, H. J. C. Numerical integration of the cartesian equations of motion of a system with constraints: molecular dynamics of n-alkanes. J. Comput. Phys. 23, 327–341 (1977).

    ADS  CAS  Google Scholar 

  46. Hanke, C. A. & Gohlke, H. In Computational Methods for Understanding Riboswitches Vol. 553 (eds. Chen, S.-J. & Burke-Aguero, D. H. B.) 163–191 (Academic Press, 2015).

  47. Roe, D. R. & Cheatham, T. E. PTRAJ and CPPTRAJ: software for processing and analysis of molecular dynamics trajectory data. J. Chem. Theory Comput. 9, 3084–3095 (2013).

    CAS  PubMed  Google Scholar 

  48. Richter, C. et al. Determination of sugar conformation in large RNA oligonucleotides from analysis of dipole–dipole cross correlated relaxation by solution NMR spectroscopy. J. Biomol. NMR 15, 241–250 (1999).

    CAS  PubMed  Google Scholar 

  49. Steger, G. Thermal denaturation of double-stranded nucleic acids: prediction of temperatures critical for gradient gel electrophoresis and polymerase chain reaction. Nucleic Acids Res. 22, 2760–2768 (1994).

    CAS  PubMed  PubMed Central  Google Scholar 

  50. Sugimoto, N. et al. Thermodynamic parameters to predict stability of RNA/DNA hybrid duplexes. Biochemistry 34, 11211–11216 (1995).

    CAS  PubMed  Google Scholar 

  51. Poland, D. Recursion relation generation of probability profiles for specific-sequence macromolecules with long-range correlations. Biopolymers 13, 1859–1871 (1974).

    CAS  PubMed  Google Scholar 

Download references


We acknowledge access to the Jülich-Düsseldorf Biomolecular NMR Center. We thank H. Arthanari, A. Boeszoermenyi and T. Viennet for valuable support regarding the 19F-NMR experiments; R. Biehl for helpful discussions; and R. Kümmerle for data acquisition at 1.1 GHz. H.G. is grateful for computational support and infrastructure provided by the "Zentrum für Informations-und Medientechnologie" (ZIM) at the Heinrich Heine University Düsseldorf and the John von Neumann Institute for Computing (NIC) (user ID: HKF7, VSK33). This work was supported by the German Research Foundation (DFG) (ET 103/2-1, ET 103/2-2, ET 103/4-1, and ET 103/5-1) to M.E., the Chemical Industry Fund (Li 196/05 to I.S. and Hoe 700080 to H.R.), the German Academic Scholarship Foundation (to H.R.), the Bayer AG (Grants4Ag program to I.S.), and the European Union’s Horizon 2020 research and innovation program under the Marie Sklodowska-Curie grant agreement number 660258 to A.V. The Center for Structural Studies is funded by the DFG (grant number 417919780).

Author information

Authors and Affiliations



J.B., J.V., H.R., A.V., C.G.W.G., C.W., M.G., H.K. and M.E. conducted the experiments and/or performed simulations, and, together with G.S., O.S., H.G. and I.S., analysed the data. NMR data acquisition and processing were carried out by J.B., A.V., M.G. and M.E. Activity assays were performed by J.V., H.R., H.K. and M.E. MD simulations were performed by C.G.W.G. EPR spectra were recorded by C.W. J.B., J.V., H.R., D.R., H.G. and M.E. wrote the manuscript. All authors designed the experiments and commented on the manuscript.

Corresponding author

Correspondence to Manuel Etzkorn.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature thanks Harald Schwalbe, Scott Silverman and the other, 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 Properties of used constructs and validation of applied modifications.

a) Table of used constructs. b) Gel electrophoretic analysis of substrate cleavage of indicated Dz variants (top, using fluorescein-labelled RNA) and of effects of 2ʹF stabilization (bottom, using GelRed staining). c) Schematic picture of 2ʹF RNA modification (left) and affected protons in [1H,1H]-TOCSY fingerprint spectra (Dz5C–RNA2ʹF, blue and Dz5C–RNA, black). The 2ʹF modification at rG0 induces CSP only in its direct proximity, i.e., at the H5 and H6 position of rU-1. While the absence of 13C enrichments in the RNA substrate impedes accurate determination of potential effects of the 2ʹF modification on the pseudorotation phase around the cleavage site48, analysis of the CSP pattern induced by the 2ʹF modification at all resolved 1H positions within the complex (d) confirms that the substrate stabilization does not alter the overall structure of the precatalytic complex. e) Nucleotide-specific ratio of peak intensities in the presence and absence of 1 mM Mg2+ for Dz5A–RNA2ʹF (red) and Dz5C–RNA2ʹF (blue). To enable a reliable comparison between different nucleotides, the changes of the cross-peak intensities of the correlation between H1ʹ and H6/H8 are shown for each Dz nucleotide. The peak disappearing in the loop region is linked to exchange processes occurring in the NMR intermediate-exchange regime, whereas the otherwise observed CSPs reveal exchange processes in the NMR fast-exchange regime. It can be concluded that Mg2+ resides longer within the catalytic loop of the 5A variant, possibly facilitating cleavage. f) Validation of labelling efficiency and cleavage capabilities of the click chemistry approach used for PRE spin labelling. A schematic of the used copper-catalysed click reaction using TEMPO-azide (2) and 5-ethynyl-2’-deoxyuridine (EdU) (1) (f, top). The EdU was used to replace one selected thymine nucleotide in the Dz sequence. To test labelling efficiency and cleavage capabilities, a FAM-azide was used, enabling direct detection via SDS-PAGE (f, bottom). To evaluate the labelling efficiency, identical amounts of a commercial FAM-labelled Dz were loaded on the indicated lane. g) Validation of activity of Dz6xF. Denaturing SDS-PAGE results of time-dependent substrate cleavage of Dz5C without and with the six 2ʹ-19F substitutions (Dz6xF). The data demonstrate that the fluorine atoms do neither affect the Dz’s cleavage activity nor Mg2+-dependency.

Extended Data Fig. 2 NMR resonance assignment of Dz5C–RNA2ʹF.

a) Extracts of a [1H,1H]-NOESY spectrum showing sequential correlations of the Dz’s sequence dG+2 to dT+9 as indicated in b). c) [1H,13C]-HSQC spectrum showing assignments for 1ʹ-CH groups (see methods for more details).

Extended Data Fig. 3 Molecular properties of the precatalytic state.

a) Presence of amino peaks as indicators of stable hydrogen-bond formation (e.g., Watson-Crick base pairing). The spectral extract of a 2D NOESY spectrum (right) containing cross correlations between cytosine H5 and the amino group as schematically highlighted (left). The data show clear signals for all base-paired cytosines present in the binding arms, whereas no signal was detected for any of the five cytosines of the Dz’s loop, indicating the absence of stable hydrogen bonding of the respective amino groups. (b,c,d) Temperature-dependent TOCSY NMR peak intensities at increasing temperatures for indicated nucleotides in either single-stranded Dz5C (b) or in the precatalytic Dz5C–RNA2ʹF complex (c,d). Deviation from a continuous intensity increase indicates denaturing of a locally stabilized structure (see Supplementary Discussion for more details). T4 shows a different behaviour than its neighbours (d). e) Comparison of [1H,1H]-TOCSY fingerprint spectra of DNAc·RNA (cyan) and DNA5C·RNA2ʹF(black). Assignments for most affected nucleotides are indicated and show that in the absence of the catalytic loop, the respective peaks around the cleavage site shift into the region characteristic for a regular duplex structure as found for the rest of the binding arms. f) Distribution of chemical shift perturbations (CSPs) induced by the presence of the catalytic loop. CSPs for all assigned protons of the binding arm region between Dz5C–RNF2ʹF and DNAc–RNA are shown as contour plot (left) or mapped on the determined structure of the precatalytic complex (right). The latter demonstrates that the regions with strong CSPs agree very well with the regions that differ from the classical A-form helix in the precatalytic complex. g) Schematic representation summarizing the NMR-based structure determination of the precatalytic complex. Indicated distance and homology restraints were combined in an ab initio structure calculation, followed by a cluster analysis and conformational filtering using RDCs, resulting in a structural ensemble (cluster I) that best resembles all experimental data (see Methods for details and Extended Data Fig. 4 for representative experimental data).

Extended Data Fig. 4 NMR-based structure determination of Dz5C–RNA2ʹF.

a,b) Experimental data and the resulting analysis of the PRE experiments. a) [1H,1H]-NOESY spectral extracts of a sample with a single TEMPO label at position T8 before (red) and after reducing the nitroxide spin label via the addition of ascorbic acid (black). b) The resulting intensity ratios between the two spectra for each resolved proton in the three samples. The marker size (in I/I0 direction) indicates the error margin (SD) obtained from the analysis of values for all resolved cross peaks of the respective proton. Only protons with at least two different resolved cross peaks were considered. c) Subset of NOE-buildup curves used for the eNOE approach. Normalized NOE intensities recorded with mixing times between 40 and 800 ms (filled circles) and buildup curves (lines) determined by eNORA32. [1H,1H]-NOESY cross peaks for the same spin system occurring above (red) or below (green) the diagonal were considered for most NOE contacts. Notably, even short and geometrically fixed distances show considerable differences in their buildup behaviour, demonstrating the necessity to incorporate cross-relaxation effects. Calculated cross-relaxation rates (σ) and extracted eNOE distance range (within 20% error margin) (blue) are indicated. d–i)19 F-based NMR experiments. d) 1D 19F-NMR spectra of Dz5C–RNA2ʹF (top) and of Dz5C variant containing six 2ʹ-19F modifications at position G−6, G−5, G2, C7, A11, and G14 (Dz6xF, bottom) in complex with RNA2ʹF. While [19F,19F]-NOESY spectra of Dz6xF–RNA2ʹF did not show any detectable peaks (data not shown), 1H-detected (e) or 19F-detected (f) [1H,19F]-HOESY spectra show a limited number of distinct cross correlations that were used for sequential resonance assignments as well as long-range distance restraints. g) H1ʹ and H2ʹ protons of the respective fluorinated ribose moieties can be identified in a [1H,1H]-NOESY spectrum via the peak splitting induced by the strong JFH coupling. h+i) To overcome sensitivity limitations of the 2D HOESY correlations, 19F-saturated and 1H-detected STD NMR was used. h) Resulting 19F-STD spectrum of Dz5C–RNA2ʹF. i) 19F-STD spectra of Dz6xF–RNA2ʹF using the indicated 19F saturation frequencies (color code refers to assignment in panel d). j-m) Characterization of residual dipolar couplings (RDCs). j) Example of RDC-induced frequency shifts in the precatalytic complex. The section shows the H5-C5 cross peak of loop position C13 in a [1H,13C]-HSQC spectra recorded in the absence (black) and presence (red) of Pf1 phage at 20 °C using 13C isotope-labelled Dz5C in complex with unlabelled RNA2ʹF. k) 1D cross section of the cross peaks shown in j) representing experimental limitations due to linewidths, peak overlap and signal-to-noise effects. l,m) Correlation plot of observed and back-calculated RDC constants for a non-matching structure of cluster III (l) and the improved correlation of the structure in cluster I (m). Although experimental limitations introduce larger error margins in the determined values (as visible by RMSD values), an RDC effect is still apparent (as visible by a considerably increased correlation).

Extended Data Fig. 5 Characterization of NMR-derived structural ensemble.

a) Results of cluster analysis after ab initio structure calculation without usage of RDCs. The 20 lowest energy structures (out of 200 calculated) were sorted according to their relative orientation of the Dz loop according to: Full winding around the RNA (cluster I), positioning downstream (cluster II) or upstream (cluster III) of the cleavage site, or other (cluster IV). Numbers below the structures refer to the number of structures in the respective fold. b) Same procedure as in a) but including RDC data. While cluster I is already the most populated in the absence of RDC restraints, it becomes by far the dominant fold after RDCs are included. In general, RDCs can be seen as one of the most reliable NMR-based probes of relative molecular orientation. As such, and albeit the remaining limitation in spectral quality and accuracy (Extended Data Fig. 4j, k), the observed effects on cluster selection appear significant. c) Overlay of the four lowest energy structures (out of 100) for each cluster (cluster I-III) after a more rigorous structural refinement and energy minimization.

Extended Data Fig. 6 Conformational plasticity of the precatalytic complex.

a) Root mean square fluctuations (RMSF) of Dz5C–RNA during MD simulations. The mean RMSF ± SEM (black area) for each nucleotide over all replicas is shown. The orange area highlights the nucleotides of the catalytic loop, and the blue areas represent nucleotides surrounding the cleavage site. b) Average structure (cartoon) from MD simulations overlaid with the structural ensemble generated from the MD trajectory (semi-transparent ribbon, with structures taken every 10 ns). c) Overlay of all seven lowest energy Dz5C–RNA2ʹF structures belonging to cluster I and obtained with loose-restraining in the ab initio structure calculation. In contrast to the ensembles shown in Extended Data Fig. 5, here, the loop region was excluded from structure alignment. It can be seen that the core region, including the binding arms and the orientation of the cleavage site, is reasonably well defined with an all-atom RMSD of 2.6 Å. On the contrary, the catalytic-loop region displays pronounced variations in the position within the different structures, providing an estimate of the allowed conformational space. Noteworthy, in direct comparison to the better-defined regions, the loop region does not display largely increased dynamics in the detected time regimes (Fig. 1f) and it still shows stacking interactions of the nucleotides with their neighbours, indicating locally defined conformations (Extended Data Figs. 9d and Supplementary Figure 1). In line with the MD-derived picture, it can hence be assumed that the Dz shows increased conformational plasticity in the loop region compared to the binding arm and, particularly, the cleavage-site region.

Extended Data Fig. 7 Metal-ion binding to the precatalytic complex.

a) Mg2+ binding induces chemical shift perturbations (CSPs) in the fingerprint spectrum of Dz5C–RNA2ʹF indicative of binding and structural rearrangements occurring in a fast-exchange regime. b) Plot of obtained nucleotide-specific KD values (black, fitted values ± standard error) as well as the magnitude of Mg2+-induced maximal CSP (red). Note that CSPs report on direct interaction and/or relayed conformational changes. Therefore, the obtained KD values can only serve as a reporter on effects (either direct or indirect) induced by Mg2+ binding. c) Mg2+-dependent cleavage activity as seen by FRET. d) Plot of atom-specific Mg2+-induced CSPs along the Dz–RNA sequence. e) Comparison between Mn2+- and Mg2+-induced activity of Dz5C. f) Mn2+ binding in the presence of excess Mg2+ does not induce CSPs but clear PRE effects allowing to distinguish between binding and structural changes. g) Different PRE rates for selected nucleotides resulting from differential distances to the Mn2+-binding site. h) Plot of atom-specific Mn2+-induced PRE rates along the Dz–RNA sequence. i) Likelihood of per-nucleotide binding of Mg2+ to the precatalytic complex during multiple microsecond-long MD simulations. j) Direct NOE cross peaks between [Co(NH3)6]3+ and Dz5C–RNA2ʹF. The dashed line indicates the frequency of the hexamine protons. In line with the observed low affinities and fast exchange rates these NOEs point to interactions of the hydration shell of the divalent metal ions with the nucleotides. The observed interactions clearly involve nucleotides of metal-ion binding region I and likely also metal-ion binding region III, confirming that central metal-ion binding sites of the precatalytic Dz–RNA complex adopt a conformation capable of accommodating the hydration shell of divalent metal ions. k) PAGE analysis of Dz5C activity in the presence of indicated concentrations of different metal ions. Interestingly, while Co2+ ions can fully activate the Dz and promote substrate cleavage, [Co(NH3)6]3+ does not mediate the Dz cleavage reaction but does inhibit Mg2+-mediated catalysis. These inhibiting effects are in line with competition for the observed same binding sides. However, the data also suggest that a hydration shell mimic does not maintain a catalytically active state. l) same as i) but, instead of cluster 1 (as used in i), structures of cluster 2–4 were used as starting point of the MD simulations. It can be seen that the MD data using cluster 1 (i) agree considerably better with the experimental data (h) than the other cluster (l), validating the cluster selection based on RDC data.

Extended Data Fig. 8 Experimental links between structure and function of the precatalytic complex.

a) The structural features of the scaffolding involve close contacts between the two binding arms, non-base pairing at position +1, and metal-ion binding in region I. b) The clear absence of respective peaks in the NOE pattern confirming flip out of dG+1. Unlike to all nucleotides with expected in-register stacking, no cross correlations of dG+1 to dG+2 is detected (which should appear along the dashed blue line). Stacking pattern for dA+6 – dT+9 is shown as positive control (green arrows). c) Effects of mutations at positions +1 and −1 on Dz5C activity confirming that position +1, unlike position −1, does not form an essential Watson-Crick base pair. d) Example of NMR data (extracts of NOESY spectra) confirming close spatial proximity of dG−5 and dG+2 after Mg2+ binding. e) Atom-specific PRE rates obtained from Mn2+ titration for DNAc–RNA (top) compared to respective nucleotides in Dz5C–RNA2ʹF (bottom). The data demonstrate that the presence of the catalytic loop dramatically changes the M2+-binding behaviour of the arms from a rather diffuse pattern (top) to the defined binding region that forms the basis of the scaffolding step (bottom). f) Top view on the precatalytic structure focusing on the cleavage-site surroundings. g) NOE-buildup rates strongly indicate syn-conformation of dA−1. The schematic model shown on top visualizes the considered inter-proton interactions with either fixed distances (H1ʹ-H2ʹ, purple, and H1ʹ-H2 ″, red) or distances strongly depending on the χ-angle (H1ʹ to indicated base proton, blue). h) NOE pattern indicative of the Mg2+-induced flip out of T4. Note the same behaviour is observed for all ten resolved inter-nucleotide correlations of T4 to its neighbours. i) Changes of cleavage activity by mutations in the 5ʹ side of the catalytic loop (metal ion-binding site II). Mutations at position 5 serve as a reference; all other mutations are variants of Dz5C. Data are presented as mean values +/- SD of triplicate experiments. j) Comparison of the cleavage activity of the variants Dz5A (black), Dz5C (blue) and Dz5G (red) in the presence of 3 mM Mg2+ (data for Dz5A and Dz5C are identical to respective conditions shown in Fig. 1b). k) Simplified schematic model highlighting the central features of the 3D structure. l-n) cw EPR characterization of Mn2+ binding. l) Double integrated EPR signal of Mn2+ in the absence (blue) and presence of 40 μM Dz5C·RNA2ʹF complexes (black). Full range and zoom into lower concentrations are shown. m,n) Fit of experimentally determined binding behaviour (black data points) for either full Mn2+ concentration range (left panels) or only the higher affinity binding sites (right panels). Data were fitted to different binding models (red curves, see Methods and Supplementary Discussion for more details on the applied data fitting). n) Three cooperative binding sites were assumed, resulting in high- and low-affinity binding sites and overall the best fit to the obtained data (left). In the right panel, a model with only one cooperative binding site was assumed, and only the shown data range was considered. The shown fit captures the key features of the binding behaviour with minimal number of parameters.

Extended Data Fig. 9 Properties of central Dz states in the catalytic cycle.

a) NMR fingerprint region of indicated states of Dz5C and its substrate in the catalytic cycle. Ellipses as shown in Fig. 1d are overlaid as a reference in each spectrum. The spectra allow to follow complex formation, effects of Na+ and Mg2+, RNA cleavage and RNA dissociation. Schematic models below spectra illustrate the respective (simplified) conditions and structural features. b) Comparison of the Dz fingerprint region before (red) and after (black) addition of RNA2ʹF. c-d) Analysis of sequential contacts occurring in NOESY spectra of free (red) and complexed Dz (black). c) Schematic of analyzed sequential contacts. d) Extract of spectral region representative of (sequential) cross correlations of the nucleotides’ base protons (H6/H8). The data indicate that a central region of the catalytic loop already adopts a (minimally structured) conformation that resembles features found for the same region after complex formation with its substrate (see Supplementary Discussion for more details). e-k) Effects of monovalent (Na+) metal ions. In the absence of (mono- and divalent) metal ions additional complex conformations are found in the NMR fingerprint spectrum (cyan arrows in e, top). These conformations disappear after addition of either 100 mM Na+ (e, middle) or 1 mM Mg2+ (e, bottom) and do not represent a catalytically relevant state (f, as visible from extrapolation of data from Extended Data Fig. 7a, grey arrow). It is tempting to speculate that this off-pathway state could represent Dz–RNA complexes in an inactive conformation characterized by hybridized binding arms lacking the additional turn of the loop around the substrate. However, other conformations are also possible. In addition to different Dz–RNA complex conformations, the spectrum in the absence of metal ions also shows the occurrence of free RNA and Dz. g) Analysis of peak volumes representative of signal from free RNA (grey), free Dz (purple), and additional complex conformation (cyan). Data were acquired in the absence of NaCl and the presence of indicated concentrations of MgCl2. A rather strong effect on complex formation (population and conformations) is found for the initial (low) Mg2+ concentrations. As visible in the spectrum recorded at 100 mM NaCl and absence of MgCl2 (e, middle), both features can also be saturated by NaCl. In other words, under conditions that are more relevant in respect to physiological ionic-strength conditions (e.g. at 100 mM NaCl) complex formation appears to not be a limiting step. However, the data strongly suggest that at low-ionic-strength conditions, strong effects in respect to complex formation and conformation (the latter likely not relevant for catalysis) can be induced by the addition of small amounts of metal ions. This observation may help to clarify previous findings involving a transition from low-ionic strength conditions. h) Results of MgCl2 NMR-titration for fingerprint region of Dz5C–RNA2ʹF in the absence (top) or presence (bottom) of 100 mM NaCl (the latter is identical to data in Extended Data Fig. 7a). i) Magnification of the behaviour of two indicated nucleotides. While the endpoints of the titrations are very similar in the absence and presence of 100 mM NaCl, the absence of NaCl leads to an extended trend with increased maximal CSPs (j). A roughly 2-fold increased KD values for Mg2+ binding is observed in the presence of NaCl. k) Correlation of the total CSPs (sum of all resolved CSPs) of the respective titration steps in the absence and presence of 100 mM NaCl. For MgCl2 concentration above 1 mM, the presence of NaCl does not have a strong effect on the Mg2+-induced CSPs. By contrast, below 1 mM MgCl2, the minimum of the correlation plot deviates from the diagonal (white arrow). The data indicate that 100 mM NaCl has a comparable effect on the Dz5C–RNA2ʹF complex structure as approx. 0.7 mM MgCl2. l) Comparison of 2D [1H-1H] TOCSY spectral extracts recorded before (black) and after full cleavage (orange). m) The nucleotides with (orange) or without (green) detected multiple conformations after cleavage highlighted on the structure of the precatalytic complex (dotted representation marks nucleotides with residual uncertainties). n) Detected signal intensities and resulting relative populations of the different states occurring in the post-catalytic complex for indicated nucleotides with increasing temperatures. o) Ratio of NMR peak intensities of the two states for best-resolved nucleotides. The values were calculated after cleavage (orange, data from spectrum in l) as well as from the real-time NMR data during the cleavage reaction (blue, data from spectrum in Extended Data Fig. 10d). The data reveal that the relative population of the respective states is stable over time, including their initial appearance during the cleavage reaction. The observed behaviour can be best explained by a stable equilibrium of slowly exchanging states (see Supplementary Discussion for more details). p) Native PAGE demonstrating cleavage of RNA substrate under NMR conditions (here, the sample after the real-time NMR experiments is shown). q) Theoretical degree of denaturation Θ at 37 °C and denaturation temperatures Tm of Dz complexes with cleaved RNA (calculated with program POLAND49 using RNA:DNA hybrid parameters50, loop parameters with σ = 1e − 351, dissociation constant β = 1e − 3, and the given complex concentrations.

Extended Data Fig. 10 Time-resolved atomic-level insights into the catalysis reaction.

a) Series of 1D 1H spectra recorded after addition of Mg2+ to pre-formed non-stabilized Dz5C–RNA complexes revealing clear time-dependent changes of the peak intensities. b) Extract of 1D spectra recorded on Dz5C–RNA before (black), directly after (blue), and at indicated time points (up to about 2 h) after Mg2+ addition. A spectrum recorded on stabilized Dz5C–RNA2ʹF, preincubated with Mg2+ for three days, serves as a reference for the Mg2+-equilibrated precatalytic complex (red). c) Contour plot of time-resolved 1D NMR data following the cleavage reaction. A magnification of data shown in (a) highlighting two peaks representing the educt (right) and product (left) state is displayed. A clear shift of the peak maximum during the initial stage of the reaction is apparent for the educt state, which is not present for the product state (dashed lines are given as guides). The direction of the observed frequency change follows the CSP induced by Mg2+ binding (visible in (b)) and therefore is in line with an increasing effective Mg2+:Dz ratio. The data strongly suggest that the post-catalytic complex has a lower affinity for Mg2+ ions than the precatalytic complex leading to an effective Mg2+ release before product release. d,e) Series of 2D [1H-1H] TOCSY spectra recorded before (black) and successively after adding 1 mM MgCl2. The acquisition time of each spectrum was 3 h. Clear peak position, peak shape, and/or intensity changes are present between the first (d, cyan) and second spectrum (e, brown). Only very weak intensity changes are present between second and third spectrum (e, orange). Indicated magnifications compare peak positions before (black) and during Mg2+-induced transition (d, cyan/blue denote positive/negative contour levels) and after full cleavage (e, orange). f) Magnification of signal for rU+6 shown in (d). The peak positions from the Mg2+ titration obtained on Dz5C–RNA2ʹF at indicated Mg2+ concentrations are shown as purple dashed lines. The regions representing the educt and product peaks in the real-time spectrum are highlighted in brown and orange. The region of the Mg2+-free state is highlighted in grey. The data reveal that the observed transition does not start from the Mg2+-free state but instead from a state that matches the frequencies of the Mg2+-equilibrated state at Mg2+ levels between 0.5 and 1 mM. g) Overlay of time-resolved experimental data for one cross peak (T8-H6/H7; positive/negative contours are denoted in blue/cyan, respectively) with best-fit simulated spectrum (red/yellow denote positive/negative contours). Characteristic features of the initial and final states are indicated. The simulations can reproduce the experimental data well. Note that data shown in panels a, b, d and g show enlarged versions of the respective data shown in Fig. 3. h) Difference between experimental and simulated real-time NMR data as a function of different rate constants applied in the simulations. Results are shown individually for each resolved atom. The minimum represents the best fit condition. The nuclei can be divided into two groups, which either show rates that are slower than the FRET rate determined under matching conditions (brown dashed line) and are plotted in blue, or faster rates (plotted in red). Respective nucleotides are mapped on the structure in Fig. 3d) 1D slice of T8(H6-H7) obtained from the indirect dimension of the 2D spectrum recorded during the cleavage reaction (d, cyan). Experimental data (black, representing state C1) is compared to peak shapes simulated using either a 2-state transition model (blue) or a 3-state transition model (red). Both simulated spectra represent the best fit for the respective model. Note that negative shoulders are better represented by the 3-state model (dotted arrows). j) Difference between the experimentally obtained data for indicated nucleotides and the simulated data as a function of the rate constant used in the simulation. While for dC+5 (grey) the overall fit is close to the experimental data (as visible by the rather low overall values of the difference), the peak is insensitive to changes in the rate constants (as visible by the low variations over the range of applied rate constants). The respective peak, therefore, is not a good sensor of the transition kinetics and was excluded from data interpretation. On the contrary, the data for C3 show a clear minimum and display good sensitivity in respect to changes in rate constants. The respective peak was therefore included in further data analysis. k) Likelihood of interactions of G14 or hexa-hydrated Mg2+ with the O2ʹ atom of rG0 or the O5ʹ atom of rU-1 at the RNA cleavage site, when in any conformation of the cleavage site (left) and when the cleavage site is in the in-line attack conformation (angle O2ʹ-P-O5ʹ: 130–180°, right). An interaction is considered present if a hetero-atom of G14 or Mg2+ is within 5 Å of the target atom. l) Effect of Mg2+ binding to the metal ion-binding site II on the frequency of structural features of the RNA cleavage reaction during MD simulations at 20 mM Mg2+. m) FRET-based activity assay in the presence of Mg2+ (red data points, 1 mM) or Mn2+ (blue data points, 0.5 mM) of Dz5C with a 6-thio-modification at either G14 (left) or at G6 (right). While the measured behaviour of 6-thio-G6 is well in line with previous reports23, the unusual behaviour of 6-thio-G14 would be in line with the acid-base mechanism shown in panel k (right, X1 = G14).

Supplementary information

Supplementary Information

This file contains Supplementary Text, Supplementary Figs. 1, 2, Supplementary Tables 1, 2 and Supplementary Equations.

Reporting Summary

Peer Review File

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Borggräfe, J., Victor, J., Rosenbach, H. et al. Time-resolved structural analysis of an RNA-cleaving DNA catalyst. Nature 601, 144–149 (2022).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

Further reading


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