See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/269338847 Visualizing phosphodiester-bond hydrolysis by an endonuclease Article in Nature Structural & Molecular Biology · December 2014 Impact Factor: 13.31 · DOI: 10.1038/nsmb.2932 · Source: PubMed CITATIONS READS 6 122 8 authors, including: Rafael Molina Stefano Stella Spanish National Research Council University of Copenhagen 31 PUBLICATIONS 276 CITATIONS 18 PUBLICATIONS 380 CITATIONS SEE PROFILE SEE PROFILE Pilar Redondo Centro Nacional de Investigaciones Oncoló… 21 PUBLICATIONS 500 CITATIONS SEE PROFILE All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Rafael Molina letting you access and read them immediately. Retrieved on: 13 July 2016 ARTICLES Visualizing phosphodiester-bond hydrolysis by an endonuclease Rafael Molina1,5, Stefano Stella1,2,5, Pilar Redondo1, Hansel Gomez3, María José Marcaida1,6, Modesto Orozco3,4, Jesús Prieto1 & Guillermo Montoya1,2 The enzymatic hydrolysis of DNA phosphodiester bonds has been widely studied, but the chemical reaction has not yet been observed. Here we follow the generation of a DNA double-strand break (DSB) by the Desulfurococcus mobilis homing endonuclease I-DmoI, trapping sequential stages of a two-metal-ion cleavage mechanism. We captured intermediates of the different catalytic steps, and this allowed us to watch the reaction by ‘freezing’ multiple states. We observed the successive entry of two metals involved in the reaction and the arrival of a third cation in a central position of the active site. This third metal ion has a crucial role, triggering the consecutive hydrolysis of the targeted phosphodiester bonds in the DNA strands and leaving its position once the DSB is generated. The multiple structures show the orchestrated conformational changes in the protein residues, nucleotides and metals during catalysis. DNA, a key molecule encoding genetic information, is subject to specificity their cuts occur in few or even single locations within many chemical processes including synthesis by polymerases, repair a genome, at DNA sequences ranging from 14 to 45 bp. Homing and editing by various factors and degradation by nucleases. All endonucleases are classified into five structural families, and sev- these processes encompass the formation and breakage of phos- eral members of the LAGLIDADG8,9 family have been engineered phodiester bonds. The discovery in the 1970s of restriction enzymes, to recognize DNA sequences other than their original targets10–13, which can specifically cleave certain DNA sequences, led to the thus suggesting that these enzymes are suitable templates to generate blooming of recombinant-DNA technologies1. However, the precise variants for specific gene targeting14–16. I-DmoI is a site-specific 17 Nature America, Inc. All rights reserved. Inc. Nature America, mechanism of DNA cleavage has yet to be established for any of intron-encoded endonuclease of the archaeon D. mobilis . This 4 these enzymes, thus hindering the development of more-efficient monomeric enzyme belongs to the LAGLIDADG homing endo- applications. Hydrolysis of phosphodiester bonds has been shown nuclease family, and it recognizes a 22-bp asymmetric target 2,3 18,19 © 201 to require divalent metal ions and to follow an SN2-type mecha- sequence, generating a staggered double-strand cut similarly nism4, which is suggested to occur in a concerted fashion leading to type II restriction enzymes9,20. Aspartate and glutamate residues to inversion of configuration at the phosphorus5,6. The products located at the end of I-DmoI LAGLIDADG motifs coordinate the of the reaction are DNA ends with a 3′-OH and a 5′-phosphate. key divalent metal ions7,9,20,21. In contrast to other LAGLIDADG Endonuclease cleavage has been analyzed by kinetic and struc- family members whose catalytic centers display three metal- tural studies. Crystal structures of different endonucleases in the binding sites22–24, only two metal sites have been observed in this apo and substrate-bound forms, to permit or prevent hydrolysis endonuclease21. of phosphodiester bonds7, have been determined in the presence Here we set out to study the catalysis of this enzyme. We devel- of different divalent metals, providing snapshots of the initial and oped a method to slow down the reaction rate in I-DmoI–DNA final catalytic states. However, to our knowledge, the course of DNA crystals, capturing the catalytic intermediates, as previously shown DSB generation by an endonuclease has never been observed to for DNA polymerases β (ref. 25) and η (ref. 26). This allowed us to date, and this has hampered the molecular understanding of this watch the course of the DNA-cleavage reaction by X-ray crystal- crucial reaction. lography. We report here the structural changes and transient ele- Homing endonucleases are a related collection of enzymes ments associated with this catalytic mechanism, providing a unique encoded either as freestanding genes within introns (as fusions with view of the reaction (Supplementary Movie 1). Furthermore, we host proteins) or as self-splicing inteins8,9. These endonucleases, like have added electronic detail to the reaction mechanism by using restriction enzymes, hydrolyze DNA, but because of their higher theoretical calculations. 1Macromolecular Crystallography Group, Structural Biology and Biocomputing Programme, Spanish National Cancer Research Centre (CNIO), Madrid, Spain. 2Macromolecular Crystallography Group, Novo Nordisk Foundation Center for Protein Research, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark. 3Joint Barcelona Computing Center (BSC)-Centre for Genomic Regulation (CRG)-Institute for Research in Biomedicine (IRB) Program in Computational Biology, Institute for Research in Biomedicine (IRB Barcelona), Barcelona, Spain. 4Departament de Bioquimica, Facultat de Biologia, University of Barcelona, Barcelona, Spain. 5These authors contributed equally to this work. 6Present address: Department of Chemistry and Biochemistry, University of Bern, Bern, Switzerland. Correspondence should be addressed to G.M. ([email protected]). Received 19 September; accepted 12 November; published online 8 December 2014; doi:10.1038/nsmb.2932 NATURE STRUCTURAL & MOLECULAR BIOLOGY ADVANCE ONLINE PUBLICATION 1 ARTICLES Figure 1 pH and temperature dependence a pH 5 6 7 8 9 c Time + – of the I-DmoI catalytic reaction. (a) pH H OH OH H O O P O 2O P P dependence of the I-DmoI DSB reaction Nicked H P O measured at 65 °C. The supercoiled and P O OH Linearized H O OH–H+ P linearized plasmid states represent the 2 OH noncleaved and cleaved states, respectively. 2+ Ground +Mn Final cleaved The I-DmoI DNA target is depicted below. Supercoiled DSB reaction state (0 h) state (b) Temperature-dependent I-DmoI cleavage I-DmoI– + – + – + – + – + activity measured at pH 6.0. Averages with Strand A coding strand 5´ 3´ s.d. of three measurements are shown. 1 2 3 4 5 6 7 8 9 10 11 12 13 d –12 –11 –10 –9 –8 –7 –6 –5 –4 –3 –2 –1 (c) Schemes of the procedure followed 3´ 5´ K120 to capture the reaction intermediates in the Strand B noncoding strand crystallized protein–DNA complex and the Q42 A116 DNA DSB reaction. (d) Detailed view of b 35 °C 40 °C 50 °C 65 °C –3C A the I-DmoI active site superimposed with the 100 2Fo – Fc electron density contoured at 1.2σ 21 80 D21 t C (PDB 2VS7) . Metal-ion coordination is shown –2C as dashed lines. The different metal-binding 2A E117 sites are alphabetically labeled (site A, B and C) 60 N129 B according to the order of cation entrance. 40 G20 % cleaved produc 3G 20 RESULTS 0 Phosphodiester-bond catalysis in crystals Noncoding strand Coding strand I-DmoI 0 5 10 15 20 25 30 35 40 45 50 D. mobilis optimal growth temperatures can Time (h) Catalytic water Water Divalent metal reach up to 90 °C in an acidic environment17, thus implying that I-DmoI’s optimal activity occurs at high tempera- at cleaving the DNA target because at this temperature the protein tures. Indeed, the protein shows a melting point at 88 °C (ref. 27). cannot overcome the activation energy of the reaction. We assayed The enzyme contained in crystals grown at 15–20 °C is inefficient I-DmoI at different pHs and found that acidic pH affected the Table 1 Data collection and refinement statistics 0 h (GS) 1 h 8 h (RS) 2 d (NS) 6 d (DSB) 8 d Data collection Space group P 21 P 21 P 21 P 21 P 21 P 21 Cell dimensions a, b, c (Å) 106.57, 70.35, 107.19, 70.85, 107.04, 70.66, 106.52, 70.15, 107.13, 70.67, 106.80, 70.41, 106.60 107.23 107.27 106.70 107.38 107.17 Nature America, Inc. All rights reserved. Inc. Nature America, 4 α, β, γ (°) 90.0, 119.89, 90.0 90.0, 119.55, 90.0 90.0, 119.58, 90.0 90.0, 119.83, 90.0 90.0, 119.62, 90.0 90.0, 119.77, 90.0 Resolution (Å) 2.70 2.60 2.73 2.30 2.55 2.30 0.07 (0.64) 0.07 (0.65) 0.11 (0.57) 0.05 (0.48) 0.05 (0.54) 0.05 (0.50) © 201 Rmerge I / σI 12.5 (2.0) 15.4 (2.4) 10.5 (2.3) 16.0 (2.0) 17.2 (2.8) 13.5 (2.7) Completeness (%) 99.7 (99.5) 97.3 (95.6) 99.9 (99.9) 96.7 (79.9) 96.6 (94.8) 95.9 (92.2) Redundancy 4.6 (4.4) 4.9 (4.8) 4.7 (4.6) 4.4 (3.0) 4.9 (4.8) 3.9 (3.8) Refinement Resolution (Å) 46.20–2.70 46.64–2.60 42.86–2.73 46.29–2.30 46.56–2.55 46.27–2.30 (2.85–2.70) (2.74–2.60) (2.85–2.73) (2.42–2.30) (2.69–2.55) (2.42–2.30) No.