Introduction, materials and methods, data availability, supplementary data, acknowledgements.
Robert Szabla, Mingyi Li, Victoria Warner, Yifeng Song, Murray Junop, DdrC, a unique DNA repair factor from D. radiodurans , senses and stabilizes DNA breaks through a novel lesion-recognition mechanism, Nucleic Acids Research , Volume 52, Issue 15, 27 August 2024, Pages 9282–9302, https://doi.org/10.1093/nar/gkae635
The bacterium Deinococcus radiodurans is known to survive high doses of DNA damaging agents. This resistance is the result of robust antioxidant systems which protect efficient DNA repair mechanisms that are unique to Deinococcus species. The protein DdrC has been identified as an important component of this repair machinery. DdrC is known to bind to DNA in vitro and has been shown to circularize and compact DNA fragments. The mechanism and biological relevance of this activity is poorly understood. Here, we show that the DdrC homodimer is a lesion-sensing protein that binds to two single-strand (ss) or double-strand (ds) breaks. The immobilization of DNA breaks in pairs consequently leads to the circularization of linear DNA and the compaction of nicked DNA. The degree of compaction is directly proportional with the number of available nicks. Previously, the structure of the DdrC homodimer was solved in an unusual asymmetric conformation. Here, we solve the structure of DdrC under different crystallographic environments and confirm that the asymmetry is an endogenous feature of DdrC. We propose a dynamic structural mechanism where the asymmetry is necessary to trap a pair of lesions. We support this model with mutant disruption and computational modeling experiments.
The bacterium Deinococcus radiodurans , along with other species of the Deinococcus genus, are distinguished for their ability to survive high doses of DNA-damaging agents, such as UV-C radiation, ionizing radiation, and desiccation ( 1 , 2 ). Several factors have been proposed to contribute to the DNA-damage resistance phenotype. Most notably is the atypically high intracellular concentration of Mn 2+ -based antioxidant species, which protects the proteome from oxidative damage both enzymatically and non-enzymatically ( 3–5 ). Shielding of the proteome enables D. radiodurans to respond rapidly to DNA damage via extremely efficient DNA repair mechanisms ( 6 , 7 ).
The response mechanism begins with the activation of the Radiation-Desiccation Response (RDR) when the bacterium senses conditions that can lead to DNA damage ( 8 , 9 ). The RDR cascade triggers the upregulation of several proteins with established connections to DNA repair including RecA, UvrA/B, GyrA/B and SSB. Among the most highly upregulated genes are five Deinococcus-unique genes: ddrA, ddrB, ddrC, ddrD and pprA (ddr, DNA-damage response; ppr, pleiotropic protein promoting DNA repair) ( 10–14 ). With the exception of DdrA, these genes do not have any identifiable sequence homologs in species outside of the Deinococcus genus. This lack of homology prompted investigations into the function of each of these upregulated genes, as they may be Deinococcus -specific DNA repair factors. Interestingly, the protein products of all five genes have been found to interact directly with DNA and effect a variety of functions related to genome maintenance. PprA is a regulator of DNA Gyrase activity, ensuring the proper resolution of various DNA topologies that arise during DNA repair ( 15–17 ). DdrA is a distant homolog of eukaryotic Rad52 ( 18–20 ). DdrB binds to single-stranded (ss) DNA and promotes accurate single-strand annealing ( 21–23 ). DdrD also binds to ss-DNA, but its function is largely unknown ( 24 ).
Multiple DNA-related behaviors have been documented for DdrC. The protein has been observed to bind to both ss- and ds-DNA ( 25 ). When bound to ss-DNA, DdrC can promote the annealing of complimentary DNA strands. When bound to ds-DNA, DdrC has been observed to circularize linear DNA, and to condense plasmids into compact shapes. Currently, it is unclear how DdrC accomplishes these various functions, nor which of these functions are biologically relevant.
In this study, we show that DdrC compacts circular dsDNA through specific interactions with ss-breaks. We demonstrate that each DdrC dimer binds to and immobilizes two ss-breaks relative to one another. DNA compaction then emerges as a consequence of trapping pairs of lesions along a DNA fragment. We also show that DdrC recognizes and immobilizes pairs of ds-breaks through an analogous break-trapping mechanism, the consequence of which is the circularization of linear DNA.
During the preparation of this manuscript, a crystal structure of DdrC was published, revealing that DdrC exists as a homodimer ( 26 ). Curiously, the structure of the DdrC homodimer is in an asymmetric conformation, where one DdrC monomer is in a different conformation from the other. This is highly unusual, as homo-oligomers of proteins typically assemble into symmetric assemblies. From the DdrC crystal structure alone, it is unclear whether the dimer asymmetry is an evolved feature or if it is a crystallographic artifact. In this study, we solved the structure of the DdrC dimer under different crystallographic environments and still observed the same asymmetric structure, indicating that dimer asymmetry is in fact an endogenous feature of DdrC.
We propose a structural mechanism where the DdrC dimer scans for and identifies DNA lesions through mechanical deformations of the DNA. In this model, the asymmetry of DdrC is an essential part of the lesion scanning process.
The ORF sequence of ddrC from D. radiodurans (Uniprot #Q9RYE6) was codon-optimized for E. coli expression, synthesized with flanking attB recombination sites, and then blunt-end subcloned into pUC57-Kan to generate a Gateway-compatible entry clone ( Supplementary Figure S12 ). Appropriate mutations were introduced by site-directed mutagenesis (SDM), while domain deletions were introduced by PCR, followed by Gateway BP cloning into pDONR-201 (Invitrogen). The DdrC variants were then cloned into publicly available E. coli expression vectors by Gateway LR cloning. A custom Gateway pDEST vector that introduces an N-terminal His 14 -mOCR fusion tag (pDEST-SHmOCR) was created for the expression of ΔNTD DdrC and deposited to Addgene (plasmid #206874). Details and creation histories for all plasmids used in this study are summarized in Supplementary Table S1 .
Selenomethionine-derivatized (SeMet) DdrC variants were expressed from their corresponding expression plasmids ( Supplementary Table S3 ) in the methionine-auxotrophic strain E. coli B834(DE3) (Novagen, Madison, WI, USA). DdrC protein was derivatized with selenium during protein expression using the M9 SeMet high-yield media kit as per the manufacturer's protocol (Medicilon, Shanghai, China). Non-derivatized DdrC variants were expressed from their corresponding expression plasmids ( Supplementary Table S3 ) in Escherichia coli BL21(DE3) by growing transformed bacteria in LB media at 37°C to an OD 600 between 0.4 and 0.7. Once the desired cell density was reached, cells were cooled on ice for 20 min, then IPTG was added to the media at a final concentration of 1 mM to stimulate protein production. The bacterial culture was then grown at 16°C for an additional 16 hours.
All downstream purification was performed at 4°C or on ice. For all DdrC variants except ΔNTD DdrC, the bacterial cells were harvested from culture by centrifugation, then washed and resuspended in Buffer A (800 mM NaCl, 5% (v/v) glycerol, 20 mM Tris, pH 8.0) at a final cell density of ∼0.1 g cells/ml. The cell suspension was then lysed by French press in the presence of protease inhibitors (1 mM Benzamidine, 1 mM PMSF, 300 nM Aprotinin, 10 μM Leupeptin, 1 μM Pepstatin A), then clarified by centrifugation at 48,000xg and filtered through a 0.45 μm PES filter. The soluble lysate was applied to a Proteindex EDTA resistant Ni-IMAC column (Marvelgent Biosciences Inc.) and washed with 40 column-volumes (CV) of Buffer A and 40 CV of 6 mM imidazole in Buffer A prior to eluting DdrC with 210 mM imidazole in Buffer A. To remove the N-terminal affinity tag, DdrC was exchanged into Buffer K (200 mM Na 2 SO 4 , 20 mM Na 3 Citrate/Citric Acid, pH 6.5) using a HiPrep 26/10 desalting column (Cytiva, LLC), then incubated with His-tagged TEV protease for 16 hours at a 1:15 TEV:DdrC molar ratio. To prevent ionic column interactions, 2 M NaCl dissolved in Buffer A was slowly added to the reaction mixture until a final NaCl concentration of 600 mM was reached. Untagged DdrC was then isolated from the reaction mixture by passing the solution over an equilibrated Ni-IMAC column, collecting the flow-through, and exchanging the protein into Buffer K. Finally, DdrC was concentrated in Buffer K using a 5 kDa MCO Vivaspin Turbo 15 PES concentrator (Sartorius AG) spun at 3000–4200×g in 10 min intervals until the desired DdrC concentration was reached ( Supplementary Table S3 ). Protein was aliquoted into single-use samples, flash-frozen, and stored at –80°C until needed. DdrC concentration was calculated from measured A 280nm values using the expected molar absorption coefficient for each DdrC variant. Purity profiles for each DdrC variant are available in Supplementary Figure S13 .
ΔNTD DdrC was purified the same way as other DdrC variants, except Buffer B (2 M NaCl, 5% (v/v) glycerol, 20 mM Tris, pH 8.0) was used in place of Buffer A and the Ni-IMAC column was washed longer, with lower concentrations of imidazole (0.45 mM for 40 CV and 0.9 mM for 40 CV). In addition, tag removal by TEV protease was done in Buffer C (50 mM KCl, 50 mM Tris pH 8.0, 0.5 mM EDTA) instead of Buffer K. Unlike the other DdrC variants, ΔNTD DdrC precipitated out of solution following tag cleavage, but the protein was readily re-solubilized by adding 2× Buffer B to the reaction mixture at a 1:1 volume ratio. Cleaved ΔNTD DdrC was then isolated, buffer exchanged, concentrated and stored the same way as all other DdrC variants.
The SNM1a nuclease was expressed and purified as previously described, then stored at 128 μM in SNM1a Storage Buffer (200 mM NaCl, 10 mM Tris pH 7.5, 5% glycerol, 0.5 mM TCEP) at –80°C ( 27 ).
Short 22-bp nicked and un-nicked dsDNA fragments were generated by annealing complimentary DNA oligonucleotides together. Oligonucleotides MJ8616, MJ8617 and MJ8618 were annealed to generate the nicked DNA ligand, while MJ8631 and MJ8618 were annealed to generate the un-nicked DNA ligand ( Supplementary Table S2 ). Similarly, MJ8381 and MJ8383 were annealed together to generate a 48-bp dsDNA fragment ( Supplementary Table S2 ). Oligonucleotides were annealed by dissolving synthesized oligos in TE buffer (10 mM Tris, pH 8.0, 1 mM EDTA) and mixing them together at a final equimolar concentration of 5 μM. The mixture was then heated to 95°C and slowly cooled to 20°C at a rate of 0.5°C/min before aliquoting and storing the annealed DNA at –20°C.
Supercoiled (RFI) and randomly nicked (RFII) ΦX174 plasmid DNA was obtained directly from New England Biolabs (NEB). The supercoiled plasmid was digested with Nt.BspQI (NEB) and FD-Eco47I (Thermo Scientific) according to the enzyme manufacturer's protocols to generate single-nicked and linear ΦX174 DNA respectively.
Supercoiled pUC19 plasmid was isolated by midi-prep (Presto™ Midi Plasmid Kit, Geneaid) from a culture of transformed E. coli DH10B (Top10) that was grown to an OD 600 of 1.8. Mutant pUC19 lacking one of three BssSI sites was generated by site-directed mutagenesis (SDM) using primers MJ8482 and MJ8483 ( Supplementary Table S2 ), then verified by Sanger sequencing. Supercoiled mutant pUC19 was then generated using the same method as WT pUC19. Both WT and mutant pUC19 were digested with nicking endonucleases that introduce exactly 1, 2, 3 or 4 ss-breaks on pUC19. The nicks in the 2-nick pUC19 were then re-sealed with T4 DNA ligase to generate relaxed, unnicked pUC19 plasmid. All pUC19 plasmids were then linearized with either SalI or SacI blunt-end restriction endonucleases. After each reaction step, the enzymes were heat-inactivated, and the reaction mixture was sufficiently diluted to overcome buffer incompatibilities of downstream enzymes. Optimized reaction conditions for each enzymatic step are provided in Supplementary Figure S3 . To generate nicked dephosphorylated pUC19 plasmid, the 3-nick variant of pUC19 was digested with rSAP (NEB) according to the manufacturer's protocol, except at a variable ratio of 0.25–4 units of rSAP per μg of pUC19. The absence of terminal 5′ phosphates was verified by digestion of 2 nM pUC19 with 200 nM SNM1A exonuclease in Buffer F (75 mM potassium acetate, 10 mM magnesium acetate, 1 mM DTT, 0.1 mg/ml BSA, 50 mM Tris/acetate, pH 7.2) for 1 h at 37°C ( Supplementary Figure S2 ).
All ΦX174 and pUC19 plasmid ligands destined for DdrC binding assays were diluted with TE buffer to a final DNA concentration of 35 ng/μl, then aliquoted and stored at –20°C. DNA concentration was calculated from A260 measurements (Nanodrop 2000c, Thermo Scientific).
All EMSA mixtures were set up and incubated at 4°C. First, DdrC was serially diluted from storage conditions with Buffer K (200 mM Na 2 SO 4 , 20 mM Na 3 Citrate/Citric Acid, pH 6.5) to 2× of the final desired protein concentration. The 2× DdrC solutions were then diluted to 1× with DNA and other buffer components to yield a final reaction mixture of DdrC in Buffer M (100 mM Na 2 SO 4 , 1 mM MgCl 2 , 0.1 mg/ml BSA, 20 mM Na 3 citrate/citric acid, pH 6.5) with either 2 nM of plasmid DNA or 100 nM of annealed DNA oligonucleotide. The EMSA reaction mixtures were then incubated for 60 min to allow DdrC to bind DNA. For the Proteinase K control, Proteinase K was then added to a final concentration of 0.2 mg/ml and incubated at 37°C for 10 min to digest DdrC. The reaction mixtures were then mixed with 6x DNA loading dye (0.025% w/v Bromophenol Blue, 30% v/v Glycerol) to a 1× final concentration. For the plasmid-based assays, 6 μl of the mixtures (10 fmol plasmid DNA) were run on a 1% w/v agarose gel in a low-pH TAE buffer (40 mM Tris, 44 mM acetic acid, 1 mM EDTA, pH 6.0) at a field strength of 4 V/cm for 2–3 h at 4°C. For the annealed oligonucleotide-based assays, 12 μl of the mixtures (1 pmol DNA) were run on a 10% native PAGE gel in the low-pH TAE buffer at a field strength of 15 V/cm for 50 min at room temperature. All gels were stained with ethidium bromide.
EMSA gels were quantified by integrating the band intensities corresponding to unbound DNA ( |${{I}_{UB}}$| ), High-affinity complex ( |${{I}_{HA}}$| ) and Low-affinity complex ( |${{I}_{LA}}$| ) in each gel lane using Image Lab software (Bio-Rad Laboratories, Inc.). The integration ranges for |${{I}_{UB}}$| , |${{I}_{HA}}$| and |${{I}_{LA}}$| were different for each DNA ligand used ( Supplementary Figure S14 ). In some cases, the signal corresponding to the UB species bleeds into the range that is integrated as the HA species due to limitations in gel resolution. This results in some proportion of the |${{I}_{UB}}$| signal being falsely integrated as |${{I}_{HA}}$| . The exact proportion of |${{I}_{UB}}$| that is falsely integrated can be calculated from the ratio of |${{I}_{HA}}$| to |${{I}_{UB}}$| signal in the 0 nM DdrC sample ( |$I_{HA}^0/I_{UB}^0$| ) since all integrated intensity in this sample must correspond to unbound DNA. The corrected |${{I}_{HA}}$| signal ( |$I_{HA}^{Corr}$| ) was therefore calculated as follows:
The LA signal must also be corrected due to a gel staining artifact where the regions near gel wells have a higher background intensity from the rest of the gel, resulting in an inflated |${{I}_{LA}}$| signal. In this case, the proportion of total DNA signal ( |${{I}_{HA}} + {{I}_{LA}} + {{I}_{UB}}$| ) being falsely integrated as |${{I}_{LA}}$| can be calculated from the ratio of |${{I}_{HA}}$| to |${{I}_{Total}}$| signal in the 0 nM DdrC sample ( |$I_{LA}^0/( {I_{HA}^0 + I_{LA}^0 + I_{UB}^0} )$| ). The corrected |${{I}_{LA}}$| signal ( |$I_{LA}^{Corr}$| ) was therefore calculated as follows:
Finally, the corrected intensities were used to calculate the proportion of DNA that is bound in the HA state and the LA state:
First, DdrC or PprA protein was serially diluted from storage conditions with Buffer K (200 mM Na 2 SO 4 , 20 mM Na 3 citrate/citric acid, pH 6.5) to 4x of the final desired protein concentration. Nicked or linear pUC19 plasmid and other buffer components were added to the protein dilutions to yield 1× DdrC or PprA with 2 nM pUC19 in a buffered background of 1× Buffer M (100 mM Na 2 SO 4 , 1 mM MgCl 2 , 0.1 mg/ml BSA, 20 mM Na 3 citrate/citric acid, pH 6.5) and 1× Buffer F (75mM potassium acetate, 10 mM magnesium acetate, 1 mM DTT, 0.1 mg/ml BSA, 50 mM Tris/acetate, pH 7.2). The mixtures were incubated at 4°C for 1 h to allow for protein–DNA complex formation. Then, either SNM1a or BglI (Fast-Digest, Thermo Scientific) were added to the mixtures at final concentrations of 200 nM for SNM1a and 0.05U/μl for BglI from 10×-concentrated stock solutions. The nuclease reactions were incubated at 37°C for 1 h. Proteinase K was then added at a final concentration of 0.2 mg/ml and incubated at 37°C for 10 min to digest the nucleases, DdrC and PprA. Finally, the reaction mixtures were loaded and run on an agarose gel as described in the ‘Electrophoretic mobility shift assay methods’ section.
To optimize buffer conditions for DdrC stability, a solution of 2x SYPRO ORANGE dye (Protein Thermal Shift™ Dye Kit, Applied Biosystems) and 175 μM WT DdrC was prepared in Buffer A (800 mM NaCl, 5% (v/v) Glycerol, 20 mM Tris, pH 8.0). The protein-dye solution was mixed at a 1:1 volume ratio with various ionic salt or buffered pH solutions in a 96-well qPCR plate (Durham Salt and pH Screens, Molecular Dimensions). All mixtures were prepared and kept at 4°C. Fluorescence in each well was monitored at 520/558 nm on a qPCR thermocycler (QuantStudio™ 3, Applied Biosystems) as the temperature was increased from 12°C to 99°C at a rate of 0.05°C/s. To identify the melting temperatures ( |${{T}_m}$| ) for each sample, signal fluctuations were first smoothed using the Protein Thermal Shift Software (Applied Biosystems), resulting in useable |$F( T )$| curves. |${{T}_m}$| values were then calculated by identifying all positive peaks of |$dF/dT$| using the SciPy signal processing library in Python ( 28 ).
To assay the relative stabilities of different DdrC variants, a solution of 2× SYPRO Orange Dye in Buffer K was mixed at a 1:1 volume ratio with 175 μM of each DdrC variant in Buffer K. The thermal melt profile of each DdrC variant was then measured and analyzed the same way as for the buffer optimization assays.
To measure the oligomeric state of DdrC, 100 μl of each DdrC variant was injected directly from storage conditions ( Supplementary Table S3 ) onto a Superdex 200 Increase 10/300 GL size-exclusion column (SEC) using the ÄKTA Pure chromatography system (Cytiva, LLC) running Buffer K at 0.5 ml/min. Absolute molecular weight was determined by SEC-coupled multi-angle light scattering analysis (SEC-MALS). The size-exclusion column was connected in-line to a Dawn HELEOS II MALS detector equipped with a 662 nm laser source and an Optilab T-rEX differential refractometer with a 658 nm LED source (Wyatt Technology). Molecular weight was calculated by Zimm plot analysis using the ASTRA software (v6.1.5.22; Wyatt Technology).
All protein crystals were grown by the hanging-drop vapor-diffusion method at 20°C. The recombinant DdrC proteins that were used for crystallization trials are summarized in Supplementary Table S3 . When screening for initial crystallization conditions, 1 μl of DdrC was mixed with 1 μl of varying precipitant solutions from different commercial screening kits. The mixed drop was suspended over 1 ml of an ammonium sulfate dehydrating solution in a sealed chamber. Conditions that yielded DdrC crystals were optimized for X-ray diffraction quality by varying: the initial DdrC concentration, the volume ratio of protein solution to precipitant solution, and the composition of the dehydrating solution. In some cases, optimized crystallization conditions were further subjected to secondary screens of additive solutions to improve crystal quality. Final crystallization conditions of the three deposited DdrC structures (PDB 7UDI, 8U0G and 8U1J) are summarized in Supplementary Table S4 .
Crystals were harvested with nylon cryo-loops and flash-frozen in liquid nitrogen. Then, crystals were mounted in a nitrogen cryo-stream at 100K during data collection. X-ray diffraction data was collected from both synchrotron and Cu K-α rotating anode radiation sources. Synchrotron data was collected at beamlines CMCF-ID and CMCF-BM of the Canadian Light Source (CLS) synchrotron, while Cu K-α data was collected on a MicroMax-007 HF generator (Rigaku Corp.). Detailed collection parameters for each of the three deposited DdrC structures are provided in Supplementary Table S5 .
All data sets were integrated and scaled using autoPROC (Global Phasing Ltd) ( 29 ). For the crystal structure corresponding to PDB 7UDI, an initial electron density map was generated by experimental SAD phasing and density modification using Phenix AutoSol ( 30 ). A model of the asymmetric unit containing two DdrC chains was built into the density map using Buccaneer, then refined iteratively in Coot and Phenix Refine ( 30–32 ). The atomic coordinates and structure factors were deposited in the Protein Data Bank under the accession 7UDI (DOI: 10.2210/pdb7UDI/pdb). Residues and sidechains with missing electron density were modeled into the 7UDI crystal structure using Rosetta Remodel with the REF2015 score function ( 33 , 34 ). The L131M/L184M mutations were also reverted to WT to yield a complete model of a FL WT DdrC dimer. This model was deposited in ModelArchive under the accession ma-nmyn0 (DOI: 10.5452/ma-nmyn0).
For the crystal structures corresponding to PDB 8U0G and 8U1J, an initial electron density map was generated by molecular replacement (MR) of the search model PDB 7UDI using Phenix Phaser ( 30 ). In the case of 8U1J, the MR search model was limited to the first 98 residues of a single chain of DdrC. The crystal structures of both 8U0G and 8U1J were built and refined the same way as 7UDI, then deposited to the PDB (DOI: 10.2210/pdb8U0G/pdb, 10.2210/pdb8U1J/pdb). Relevant data processing and model refinement statistics are available in Table 3 .
Inter- and intra-molecular contacts within the DdrC dimer were identified from the filled 7UDI crystal structure by automated algorithms in PyMOL (Schrödinger, LLC). To identify possible domain boundaries by structural homology, the DdrC structure was queried for structural similarity against the entire PDB databank using the DALI protein structure comparison server ( 35 ). The DALI results were then analyzed using DALIview (DOI: 10.5281/zenodo.8435478) to reveal structurally similar domain families.
Axes of symmetry were identified within the DdrC dimer by calculating the midpoint positions in 3D space between every atom in chain A and its analogous atom in chain B. The list of 3D midpoints was grouped according to DdrC domains, then subjected to Principal Component Analysis (PCA), yielding a list of PCA eigenvectors for each DdrC domain. The longest eigenvector for each domain corresponds to a fitted C2 axis of symmetry when plotted in 3D space.
To identify the source of DdrC dimer asymmetry, the angular differences between chain A and chain B backbone torsion angles were calculated for all phi ( |$\varphi$| ) and psi ( |$\psi$| ) angles. The absolute angle differences at each residue position ( |$| {\Delta \varphi } | + | {\Delta \psi } |$| ) were mapped onto the structure in PyMOL according to a color gradient.
Electrostatic surface potentials were calculated for experimental and theoretical DdrC structures using the APBS software suite ( 36 ).
To generate an atomic model of DdrC in complex with un-nicked intact dsDNA, a DNA duplex of arbitrary sequence and length was docked onto the filled 7UDI crystal structure using the rigid protein/flexible DNA algorithm, Paradock ( 37 ). The top-scoring output model from Paradock was then refined against the Rosetta REF2015 score function through 1000 stochastic repetitions of the Rosetta relax protocol ( 34 , 38 ). The top-scoring model was deposited to ModelArchive under the accession ma-urph3 (DOI: 10.5452/ma-urph3) and used for downstream analysis.
A model of DdrC in complex with a 21 bp nicked DNA duplex was predicted using the RosettaFoldNA neural net model (v0.1) from the amino acid sequence of DdrC and the nucleotide sequence of 3 complimentary DNA oligonucleotides (CGTCATCACCGAAACGCGCGA, TCGCGCGTTTCGG and TGATGACG) ( 39 ). The top-scoring output model was then refined using the same method as for the un-nicked DNA model, except with an explicit C2 axis of symmetry. The top-scoring model was deposited to ModelArchive under the accession ma-50nj9 (DOI: 10.5452/ma-50nj9).
Finally, a model of DdrC in complex with a terminal dsDNA end was generated from the nicked dsDNA model by removing nucleotides that are downstream from the nick. The resulting model was refined and deposited to ModelArchive under the accession ma-otnza (DOI: 10.5452/ma-otnza).
The native DNA sequence of D. radiodurans ddrC was synthesized in the form of either WT, NTD-mut or CTD-mut variants together with the Deinococcal constitutive promoter PDR_1261 ( Supplementary Figure S18 ) ( 40 ). The synthesized DNA fragments were then subcloned into pRad1 at XhoI / XbaI restriction sites to generate a series of ddrC complementation plasmids ( 41 ). Plasmid details are summarized in Supplementary Table S1 . D. radiodurans R1 strains harboring ΔuvsEΩhygro and ΔuvsEΩhygroΔddrCΩkan genomic deletions were obtained from Dr. Fabrice Confalonieri and transformed with plasmid as previously described ( 25 , 42 ). Transformants were selected and cultured in the presence of 50 μg/ml Hygromycin-B, 6 μg/ml Kanamycin or 3 μg/ml Chloramphenicol, as appropriate. Each transformant was cultured in liquid 2×TGY media at 32°C to an OD 600 of 1.0, then serially diluted, spot plated, and exposed to a UV-C light source (254 nm) at a fixed distance for varying lengths of time. UV exposure dose was calculated from radiometer dose rate measurements. The UV-treated TGY-Agar plates were incubated at 30°C for 60 h, then imaged and analyzed. The surviving fraction of D. radiodurans was measured in triplicate by comparing the CFU counts at each UV-C dose to unirradiated bacteria.
When DdrC is incubated with supercoiled, relaxed, or linear forms of ΦX174 plasmid dsDNA, the DNA mobility is shifted into the well at high DdrC concentrations (Figure 1A – D ). This suggests the formation of a large intermolecular complex. It is unclear whether this species is of any biological relevance. At lower DdrC concentrations (<300 nM DdrC per nM DNA), DdrC forms smaller complexes with DNA, as the bound species migrates into the gel.
Gel motility shifting of dsDNA upon DdrC binding. 2 nM of ( A ) supercoiled, ( B ) linear, ( C ) randomly nicked and ( D ) single-nicked ΦX174 dsDNA plasmids incubated with DdrC at varying concentrations. ( E ) Addition of Proteinase K to a pre-formed DdrC-DNA complex with randomly nicked ΦX174.
Interestingly, the electrophoretic mobility of this smaller species is shifted differentially by DdrC depending on the starting plasmid topology. Both linear and supercoiled ΦX174 are shifted to a discrete species in the presence of DdrC (Figure 1A , B ), but, the shift in mobility of supercoiled DNA is progressive, whereas that of linear DNA is sudden ( Supplementary Figure S1 ). In other words, supercoiled plasmid appears to have multiple DdrC binding sites, whereas linear plasmid only has one.
The most surprising binding behaviour across the different ΦX174 isomers is the observation that relaxed ΦX174 results in an increase of plasmid mobility. This is unexpected, as DNA binding proteins typically impede the mobility of their DNA binding partners on a gel, rather than increase it. An increase in DNA mobility can only occur in three situations: (i) the DNA/protein complex is more negatively charged than the DNA on its own, (ii) the DNA sequence is physically shortened by nuclease digestion or (iii) the DNA undergoes topological changes which lowers its radius of gyration.
It is highly unlikely that the increased DNA mobility is the result of a more negative net charge upon complexation with DdrC, as DdrC has a theoretical p I of 9.7 and is expected to have a net charge of +5 under Buffer M conditions. To examine whether the change in mobility is due to nuclease activity, Proteinase K was added to the DNA following incubation with DdrC (Figure 1E ). The addition of Proteinase K to a pre-formed DdrC-DNA complex restores the mobility of the plasmid to its unbound state, demonstrating that the increase in mobility is not a result of nuclease degradation. It has been shown previously by TEM that DdrC induces DNA compaction of relaxed plasmid DNA ( 25 ). We propose that the fast-moving species formed upon the addition of DdrC is in fact compacted plasmid. Interestingly, compaction of ΦX174 plasmid by DdrC is observed to a much greater degree when the plasmid contains many randomly generated single-strand breaks compared to plasmid harboring a single enzymatically-produced nick (Figure 1C , D ). So, plasmid compaction by DdrC appears to be dependent on the presence of nicks.
Since plasmid compaction seems to be dependent on the presence of ss-breaks, it seems very likely that DdrC is recognizing and binding directly to DNA nicks. We tested this assumption by assaying the binding characteristics of DdrC to a short 22-mer DNA duplex with and without an internal ss-break (Figure 2A ). As expected, a discrete band shift was observed only when a nick was present. This result demonstrates that DdrC recognizes and binds to ss-breaks directly. Furthermore, it appears that DdrC is stabilizing the nicked DNA duplex as the DNA transforms from a smeared, diffuse band in the unbound state to a sharp, discrete band in the bound state. Stoichiometrically, the DNA band was nearly fully shifted at a ratio of 2 DdrC monomers per nick, suggesting that DdrC binds to the DNA as a dimer. But this does not explain the mechanism by which DdrC compacts plasmids. Binding of DdrC to the 22-mer duplex results in an upwards band shift, as opposed to a downwards shift, as observed with ΦX174 plasmid harboring multiple nicks. This is likely because the 22-mer DNA fragment is too short to become compacted. Also, it appears that the degree of compaction may be dependent on the number of nicks available on the DNA.
Characterization of the interactions between DdrC and single-strand breaks. ( A ) DdrC-induced motility shift of a 22 bp dsDNA fragment at 100 nM with an internal nick that is either present or absent. ( B ) DdrC-induced motility shift of pUC19 plasmid at 2 nM that has been pretreated with either Nt. BspQI or Nb.BssSI nicking endonucleases. ( C ) SNM1a exonuclease or BglI endonuclease digestion of nicked pUC19-DdrC complexes at varying DdrC concentrations. ( D ) DdrC-induced motility shift of nicked pUC19 that has been dephosphorylated with rSAP. ( E ) Motility shift of variably-nicked pUC19 plasmid following incubation with 75 nM of DdrC. Terminal 5′ phosphates at single-strand break sites are depicted with red circles. Relative positions of nicks on plasmid schematics are to scale. ( F ) Proposed interpretation of plasmid compaction assay.
Since multiple random nicks on ΦX174 plasmid yields a smear of compacted species, while a single nick yields almost no compaction, we hypothesized that a plasmid harboring a discrete number nicks would result in compaction to a discrete species. To test this hypothesis, DdrC was incubated with pUC19 that has been relaxed from its supercoiled state using Nt.BspQI or Nb.BssSI nicking endonucleases, which introduce exactly one or three nicks on pUC19, respectively (Figure 2B ). Like with the ΦX174 plasmid, no compaction was observed when there was only one nick present. But when three nicks were present, DdrC compacted the pUC19 plasmid to a single, discrete species, as expected. This result demonstrates that the mechanism of plasmid compaction by DdrC is dependent on the number of available nicks.
Given that DdrC clearly binds to a short, nicked duplex (Figure 2A ), it is likely that the 1-nick pUC19 plasmid is also being bound by DdrC, despite the lack of a visible band shift. We hypothesize that a single nick is sufficient for DNA binding, but insufficient for DNA compaction, so the plasmid remains in a circular, relaxed topology when bound by DdrC. To verify that DdrC binds to the ss-break on the 1-nick pUC19 plasmid, the plasmid was incubated with varying concentrations of DdrC, then treated with either an SNM1a, or BglI nucleases (Figure 2C ). SNM1a is a 5′→3′ exonuclease that initiates digestion from nick sites ( 27 , 43 ), while BglI is a is sequence-specific endonuclease that happens to digest pUC19 at two remote sites that are distant from the three engineered nicks. Under our assay conditions, DdrC protects DNA from SNM1a exo-digestion, but not BglI endo-digestion. In contrast, the DNA-binding protein PprA from D. radiodurans , which is known to form protein filaments, inhibits both endo- and exo-nuclease activity under the same conditions ( Supplementary Figure S15 ) ( 44 , 45 ). These findings demonstrate that DdrC binds to dsDNA directly at the site of the ss-break site and nowhere else on the plasmid. As such, the lack of a downwards band shift in the 1-nick pUC19 plasmid is not due to a lack of DNA binding, but because a single nick is insufficient for the DdrC compaction mechanism.
To further investigate the mechanism of nick recognition, we examined the requirement of a 5′ terminal phosphate for DdrC compaction, since the presence of a free 5′ phosphate chemically differentiates nicked from unnicked dsDNA, and dedicated phosphate recognition sites have been identified in a large subset of proteins in the PDB ( 46 ). First, terminal 5′ phosphates were removed from the 3-nick pUC19 substrate using varying amounts of Shrimp Alkaline Phosphatase (rSAP). SNM1a, a 5′ phosphate-dependent exonuclease, was used to probe for available 5′ phosphates on the rSAP-treated pUC19 samples. No SNM1a digestion was detected when treating pUC19 with 4 U rSAP/μg DNA, indicating that there are no available 5′ phosphates in the 4 U/μg population of nicked pUC19 ( Supplementary Figure S2 ). This sample of pUC19 was then incubated with DdrC (Figure 2D ). Treatment of 3-nick pUC19 with rSAP has no effect on subsequent DNA compaction by DdrC. So the mechanism of nick binding and plasmid compaction by DdrC must involve a mechanism that does not rely on interactions with a terminal 5' phosphate. Many DNA lesion sensing proteins evolved to exploit the altered mechanical properties of DNA at the site of lesions ( 47 ). We hypothesize that DdrC may rely on a similar mechanical sensing mechanism since a DNA duplex would have higher conformational freedom at the site of a ss-break.
The data presented so far strongly implies that there is a relationship between the number of available of nicks on a plasmid and the degree to which that plasmid becomes compacted by DdrC. We have observed that plasmid with a single nick does not undergo significant compaction by DdrC, whereas a plasmid with exactly 3 nicks is compacted to a single species, and a randomly nicked plasmid is compacted to an even greater degree with a large distribution in gel mobility.
To examine the relationship between DNA compaction and the number of available single-strand breaks, a series of variably nicked pUC19 plasmids were generated. The sequence of pUC19 allows for the addition of evenly spaced nicks using commercially available nicking endonucleases. This was possible because the pUC19 sequence contains 1 Nt.BspQI site, 3 Nb.BssSI sites and 4 Nb.BstNBI sites. Then, a variant of pUC19 was generated via mutagenesis to contain only two Nb.BssSI sites to prepare a plasmid with exactly 2 nick sites,.
Nt.BspQI, Nb.BssSI and Nb.BstNBI endonucleases were then used to prepare the series of topologically relaxed pUC19 plasmids containing 0, 1, 2, 3 and 4 single-strand breaks ( Supplementary Figure S3 ). This series of nicked pUC19 plasmids were then incubated with DdrC (Figure 2E ). As expected, it was observed that DdrC induces compaction of the nicked plasmids to discrete species and that the degree of compaction increases with the number of available nicks.
To measure the degree of compaction, a pUC19 ‘compaction marker’ was prepared and run on the gel alongside the DdrC-compacted species (Figure 2D ). The compaction marker was generated by nicking and subsequently re-ligating pUC19 plasmid, yielding a mixture of at least five supercoiled topoisomers of pUC19. This was possible as circular, nicked DNA can freely rotate around the strand opposite from the nick due to random thermal motion, leading to spontaneous over- or under-winding of the DNA duplex. Ligation of the nicks then traps the molecules in discrete supercoiled states and produces a set of identical plasmids that differ only in their topological linking number (Δ L k ) by positive or negative integer values ( 48 , 49 ). The distribution of the Δ L k values is Gaussian and can be altered by varying the temperature during ligation ( 48 , 49 ). This plasmid mixture effectively acts as a quantized DNA marker for pUC19 supercoiling with linking numbers (Δ L k ) between 0 and 4. The mobility of the DdrC-compacted species was then measured relative to the compaction marker (Figure 2D ). Interestingly, the mobility of the DdrC-compacted species also appears to be ‘quantized’ as they only seem to take on values corresponding to specific pUC19 topoisomers in the marker (Table 1 ). In short, a single nick leads to an apparent Δ L k of 0. Two and three nicks both result in an apparent Δ L k of 1, and four nicks results in a Δ L k value of 2.
Apparent linking numbers of DdrC-bound pUC19 plasmids—circular
DNA sample . | Unbound (–DdrC) . | Bound (+DdrC) . | ||
---|---|---|---|---|
. | Normalized mobility . | Linking number (|$\Delta {{L}_k}$|) . | Normalized mobility . | Linking number (|$\Delta {{L}_k}$|) . |
Topo marker | 0.000 | 0.00 | 0.000 | 0.00 |
0.219 | 1.00 | 0.262 | 1.00 | |
0.499 | 2.00 | 0.538 | 2.00 | |
0.765 | 3.00 | 0.800 | 3.00 | |
1.000 | 4.00 | 1.000 | 4.00 | |
0-nick | -0.063 | -0.19 | 0.091 | 0.32 |
0.157 | 0.67 | 0.245 | 0.92 | |
0.452 | 1.82 | 0.538 | 2.07 | |
0.702 | 2.81 | 0.769 | 2.98 | |
0.937 | 3.73 | 0.938 | 3.64 | |
1-nick | –0.110 | –0.38 | 0.029 | 0.07 |
2-nick | –0.078 | –0.26 | –0.017 | –0.11 |
0.262 | 0.99 | |||
3-nick | –0.078 | –0.26 | 0.262 | 0.99 |
4-nick | –0.063 | –0.19 | 0.493 | 1.89 |
DNA sample . | Unbound (–DdrC) . | Bound (+DdrC) . | ||
---|---|---|---|---|
. | Normalized mobility . | Linking number (|$\Delta {{L}_k}$|) . | Normalized mobility . | Linking number (|$\Delta {{L}_k}$|) . |
Topo marker | 0.000 | 0.00 | 0.000 | 0.00 |
0.219 | 1.00 | 0.262 | 1.00 | |
0.499 | 2.00 | 0.538 | 2.00 | |
0.765 | 3.00 | 0.800 | 3.00 | |
1.000 | 4.00 | 1.000 | 4.00 | |
0-nick | -0.063 | -0.19 | 0.091 | 0.32 |
0.157 | 0.67 | 0.245 | 0.92 | |
0.452 | 1.82 | 0.538 | 2.07 | |
0.702 | 2.81 | 0.769 | 2.98 | |
0.937 | 3.73 | 0.938 | 3.64 | |
1-nick | –0.110 | –0.38 | 0.029 | 0.07 |
2-nick | –0.078 | –0.26 | –0.017 | –0.11 |
0.262 | 0.99 | |||
3-nick | –0.078 | –0.26 | 0.262 | 0.99 |
4-nick | –0.063 | –0.19 | 0.493 | 1.89 |
a Mobility distances are measured relative to the Δ L k = 0 and Δ L k = 4 samples in the topo marker.
b Linking numbers are calculated by linear interpolation from Topo marker standards.
Each of the five species in the compaction marker is expected to have a set number of writhe points, which are figure-8 structures that form spontaneously to resolve the torsional strain of supercoiling ( 50 ). These differences in plasmid shape account for the differences in gel mobility. Since the mobility of the DdrC-compacted plasmids appear to match the mobility of specific pUC19 topoisomers, we hypothesize that the DNA structures formed upon DdrC binding are similar to the writhe point structures that arise in supercoiled plasmids. It is also apparent that DdrC-mediated compaction only occurs when there are two or more nicks available on the plasmid, after which point, the apparent linking number of the bound pUC19 increases by one with every two additional nicks. These observations offer insight into the mechanism of DNA compaction by DdrC (Figure 2F ). Simultaneous binding of two distal nick sites by a single protein would produce a topological constraint on DNA that looks like a single DNA writhe point by agarose gel electrophoresis. Since the apparent linking number of the bound pUC19 increases by one with every two additional nicks, we can conclude that each unit of DdrC recognizes and binds to two nicks on the plasmid. As such, these data strongly suggest that the mechanism of DdrC plasmid compaction occurs through bridging two distal nick sites into close spatial proximity.
Although we did not detect any binding to unnicked, relaxed pUC19, we did observe significant binding of DdrC to supercoiled pUC19 ( Supplementary Figure S4 ). The same behavior was observed with ΦX174 plasmid ( Supplementary Figure S1 ). Unlike binding to other dsDNA substrates, the degree of band shifting appears to be progressive in the case of supercoiled plasmid. In other words, a higher concentration of DdrC yields a more prominent shift, up to a saturation point. This correlation indicates that each supercoiled plasmid can be bound by multiple molecules of DdrC. Supercoiled plasmid extracted from E. coli is expected to contain multiple writhe points. It is very likely that DdrC is binding to the writhe points of supercoiled plasmids, as these DNA structures may be the same structures that are induced when DdrC bridges two nicks during plasmid compaction. It follows then that the topology of DdrC-compacted DNA resembles the topology of supercoiled DNA; however, it is unclear whether DdrC actively supercoils DNA or if it simply induces a writhe point-like structure without over- or under-winding the DNA duplex.
The binding of DdrC to linear ΦX174 plasmid results in an upwards band shift to a single, discrete position (Figure 1B ). This result indicates that the protein–DNA complex is assembled from a fixed stoichiometric ratio of DdrC to DNA. Given that DdrC binds directly to single-strand breaks, it is highly likely that DdrC binds linear DNA fragments at double-strand break sites, which would explain why the complex has a discrete stoichiometry of DdrC to plasmid.
To characterize how DdrC interacts with double-strand breaks, we assessed whether DdrC has a preference for overhang type at the terminal ends of linear DNA. In this assay, one ds-break was introduced into pUC19 plasmid harboring either 0, 2 or 4 nt overhangs (5′) using SmaI, NdeI and SalI endonucleases, respectively. The varying overhang DNA fragments were then incubated with varying concentrations of DdrC and the shift profiles were visualized on a gel (Figure 3A ). It is clear that DdrC has a significantly higher affinity for blunt ds-breaks when compared to overhangs: ∼2-fold higher when compared to a 2 nt overhang, and ∼7-fold when compared to a 4 nt overhang.
Characterization of the interactions between DdrC and double-strand breaks. ( A ) DdrC-induced motility shift of pUC19 plasmid at 2 nM that has been pre-treated with three different endonucleases: SmaI, NdeI and SalI. ( B ) Motility shift of blunt-end linear pUC19 compared to unbound, circular pUC19 ( C ) Motility shift of blunt-end linear pUC19 plasmids following incubation with 75 nM of DdrC. Linearized pUC19 plasmids were relaxed with nicking endonucleases to harbor a specific number of nicks. Nick sites are represented with red triangles and their relative positions on the linear plasmid schematics are to scale. ( D ) Proposed model of linear plasmid circularization and compaction by DdrC.
Despite differences in affinity, the shifted DNA migrates to the same discrete position regardless of overhang type. Interestingly, the gel migration position of bound, linear pUC19 appears to be the same as unbound, relaxed circular pUC19 (Figure 3B ). This suggests that the topology of the two species is the same. In other words, DdrC appears to be circularizing linear DNA. If this is the case, then we would expect linear pUC19 harboring single-strand breaks to become compacted in the same way as circular nicked pUC19.
To test this hypothesis, we generated blunt-end, linear pUC19 harboring 0, 1, 2, 3 and 4 nicks ( Supplementary Figure S3 ). This series of linear plasmids was then incubated with DdrC (Figure 3C ). As expected, compaction is observed with the nick-harboring DNA. As with the circular plasmids, the degree of compaction scales with the number of available nicks on the plasmid (Table 2 ). This result strongly suggests that DdrC circularizes linear DNA via bridging of ds-breaks, then circularized plasmid can become compacted via bridging of ss-breaks (Figure 3D ). The nick-bridging model of plasmid compaction by DdrC suggests that each functional unit of DdrC has two DNA binding sites. Each binding site can recognize and bind to either a ss-break or ds-break.
Apparent linking numbers of DdrC-bound pUC19 plasmids—linear
DNA sample . | Unbound (–DdrC) . | Bound (+DdrC) . | ||
---|---|---|---|---|
. | Normalized mobility . | Linking number (|$\Delta {{L}_k}$|) . | Normalized mobility . | Linking number (|$\Delta {{L}_k}$|) . |
Topo marker | 0.000 | 0.00 | 0.000 | 0.00 |
0.193 | 1.00 | 0.164 | 1.00 | |
0.496 | 2.00 | 0.474 | 2.00 | |
0.773 | 3.00 | 0.759 | 3.00 | |
1.000 | 4.00 | 1.000 | 4.00 | |
0-nick | 0.210 | 0.84 | 0.009 | 0.04 |
1-nick | 0.210 | 0.84 | 0.198 | 0.81 |
2-nick | 0.176 | 0.71 | 0.284 | 1.16 |
3-nick | 0.244 | 0.97 | 0.336 | 1.37 |
4-nick | 0.261 | 1.04 | 0.491 | 2.00 |
0.647 | 2.63 |
DNA sample . | Unbound (–DdrC) . | Bound (+DdrC) . | ||
---|---|---|---|---|
. | Normalized mobility . | Linking number (|$\Delta {{L}_k}$|) . | Normalized mobility . | Linking number (|$\Delta {{L}_k}$|) . |
Topo marker | 0.000 | 0.00 | 0.000 | 0.00 |
0.193 | 1.00 | 0.164 | 1.00 | |
0.496 | 2.00 | 0.474 | 2.00 | |
0.773 | 3.00 | 0.759 | 3.00 | |
1.000 | 4.00 | 1.000 | 4.00 | |
0-nick | 0.210 | 0.84 | 0.009 | 0.04 |
1-nick | 0.210 | 0.84 | 0.198 | 0.81 |
2-nick | 0.176 | 0.71 | 0.284 | 1.16 |
3-nick | 0.244 | 0.97 | 0.336 | 1.37 |
4-nick | 0.261 | 1.04 | 0.491 | 2.00 |
0.647 | 2.63 |
Previously, it has been reported that DdrC has a higher affinity for ssDNA than dsDNA at a fragment size of 67-mer ( 25 ). In both cases, the DNA fragments were shifted to a discrete position on the gel, indicating that the complex is composed of a fixed stoichiometric ratio of DdrC to DNA. A fixed ratio suggests that DdrC is binding to both DNA fragments at the termini. We hypothesize that the apparent preference for ssDNA may be a symptom of selective DNA circularization by DdrC. Since the persistence length of dsDNA is about 50 nm (150 bp), while that of ssDNA is about 0.75nm (<5 nt), then a DNA length of 67-mer may be sufficient for circularization of ssDNA, but not dsDNA ( 51 ). Consequently, a 67-mer fragment of ssDNA may be able to contact both binding sites on DdrC, while a dsDNA duplex of the same length would only contact one. The ability of one ligand to contact more binding sites should correspond with a higher apparent binding affinity in a gel shift assay. To test this hypothesis, the experiment by Bouthier de la Tour et al. was repeated using shorter 48-mer ssDNA and dsDNA ligands ( Supplementary Figure S16 ). Under these conditions, we do not observe a preference for either dsDNA or ssDNA, suggesting that in both cases, the DNA is only contacting a single binding site on DdrC. We hypothesize that the 48-mer ssDNA is either too short, too rigid, or contains secondary structures that prevent it from contacting both binding sites on DdrC.
To gain insight on the molecular mechanism of DdrC nick detection, we solved the structure of DdrC by X-ray crystallography. Crystals of selenomethionyl (SeMet)-derivatized DdrC were grown and diffracted at the X-ray wavelength corresponding to the absorption edge of selenium. Although the resulting diffraction data was of high quality and resolution (Table 3 ), the anomalous signal did not have sufficient phasing power to recover the phases of measured structure factors. We suspect that too few methionine residues are to blame for the weak anomalous signal, as D. radiodurans DdrC only contains one internal methionine per monomer, and so a SeMet-derivatized DdrC crystal could only contain one structured selenium atom for every 231 native residues.
X-ray data collection and processing statistics
PDB accession code . | 7UDI . | 8U0G . | 8U1J . |
---|---|---|---|
Crystalized protein | DdrC, FL, L131M/L184M | DdrC FL, WT | DdrC 1–98, WT |
Heavy atom derivatization | SeMet | Native | SeMet |
Space group | 4 | 3 2 1 | 4 2 2 |
Cell dimensions: (Å) | 66.70, 66.70, 129.58 | 111.04, 111.04, 101.49 | 73.38, 73.38, 110.18 |
Cell dimensions: (°) | 90, 90, 90 | 90, 90, 120 | 90, 90, 90 |
Resolution (Å) | 2.239 (2.278) | 4.277 (4.351) | 2.972 (3.023) |
0.055 (0.706) | 0.182 (0.987) | 0.060 (0.479) | |
19.1 (2.3) | 9.9 (2.9) | 33.7 (5.4) | |
Completeness (%) | 96.8 (99.9) | 100.0 (100.0) | 100.0 (100.0) |
Redundancy | 10.1 (5.2) | 8.9 (9.1) | 20.9 (20.2) |
No. reflections: total | 332 975 (17 420) | 46 579 (2199) | 69 588 (3311) |
No. reflections: unique | 32 071 (1600) | 5236 (242) | 3335 (164) |
/ | 0.231 / 0.250 | 0.252 / 0.346 | 0.225 / 0.257 |
No. atoms | 3293 | 6844 | 748 |
Protein | 3233 | 6844 | 748 |
Ion | 20 | 0 | 0 |
Water | 40 | 0 | 0 |
-factors | 66.54 | 237.80 | 72.09 |
Protein | 66.51 | 237.80 | 72.09 |
Ion | 91.62 | 0 | 0 |
Water | 56.00 | 0 | 0 |
Bond length RMSD (Å) | 0.021 | 0.006 | 0.020 |
Bond angle RMSD (°) | 1.751 | 0.978 | 1.662 |
PDB accession code . | 7UDI . | 8U0G . | 8U1J . |
---|---|---|---|
Crystalized protein | DdrC, FL, L131M/L184M | DdrC FL, WT | DdrC 1–98, WT |
Heavy atom derivatization | SeMet | Native | SeMet |
Space group | 4 | 3 2 1 | 4 2 2 |
Cell dimensions: (Å) | 66.70, 66.70, 129.58 | 111.04, 111.04, 101.49 | 73.38, 73.38, 110.18 |
Cell dimensions: (°) | 90, 90, 90 | 90, 90, 120 | 90, 90, 90 |
Resolution (Å) | 2.239 (2.278) | 4.277 (4.351) | 2.972 (3.023) |
0.055 (0.706) | 0.182 (0.987) | 0.060 (0.479) | |
19.1 (2.3) | 9.9 (2.9) | 33.7 (5.4) | |
Completeness (%) | 96.8 (99.9) | 100.0 (100.0) | 100.0 (100.0) |
Redundancy | 10.1 (5.2) | 8.9 (9.1) | 20.9 (20.2) |
No. reflections: total | 332 975 (17 420) | 46 579 (2199) | 69 588 (3311) |
No. reflections: unique | 32 071 (1600) | 5236 (242) | 3335 (164) |
/ | 0.231 / 0.250 | 0.252 / 0.346 | 0.225 / 0.257 |
No. atoms | 3293 | 6844 | 748 |
Protein | 3233 | 6844 | 748 |
Ion | 20 | 0 | 0 |
Water | 40 | 0 | 0 |
-factors | 66.54 | 237.80 | 72.09 |
Protein | 66.51 | 237.80 | 72.09 |
Ion | 91.62 | 0 | 0 |
Water | 56.00 | 0 | 0 |
Bond length RMSD (Å) | 0.021 | 0.006 | 0.020 |
Bond angle RMSD (°) | 1.751 | 0.978 | 1.662 |
a Data collection statistics for high-resolution shell indicated in parentheses.
b Atoms with non-zero occupancy.
To overcome this obstacle, we introduced additional SeMet residues into DdrC at structured positions where SeMet substitution seems structurally and functionally safe. These two positions, Leu-131 and Leu-184, are frequently substituted to methionine in other close DdrC homologs that have highly conserved local sequence, and so L131M and L184M substitutions should have minimal effect on local and global folding in D. radiodurans DdrC ( Supplementary Figure S5 ). Crystals of SeMet-derivatized L131M/L184M DdrC were then grown, diffracted, and successfully phased, allowing for the solution and refinement of the crystal structure. This structure was deposited in the PDB under the accession code 7UDI (Figure 4A , Table 3 ). The structure of 7UDI was then used to phase diffraction data from the WT DdrC crystal by molecular replacement (MR). This structure was deposited in the PDB under accession code 8U0G ( Supplementary Figure S6 , Table 3 ).
Structural characterization of DdrC domains. ( A ) Crystal structure of the full-length DdrC homodimer colored according to predicted domain boundaries. Residues with missing electron density were modeled using Rosetta remodel and are represented with a dashed backbone. All inter- and intra-molecular sidechain interactions within the dimer are represented on the 1D domain map. ( B ) Differential scanning fluorimetry profiles of 3 truncated DdrC variants. The T m value corresponding to each dF/dT peak is indicated with an arrow. ( C ) Crystal structure of a proteolytically-degraded sample of DdrC. The integrity of the protein in the source crystallization drop was verified by SDS-PAGE pre- and post- crystallization and was compared to the integrity of the protein in the FL DdrC crystallization condition. The expected positions of possible DdrC species are indicated with arrows. ( D ) SEC-MALS analysis of the oligomeric state of three truncated DdrC variants.
Interestingly, both crystal structures had different lattice symmetries and crystal contacts. Despite the difference in lattice contacts, one common interaction interface persisted between the two structures (Figure 4A ). Conservation of this protein-protein contact strongly suggests that this is a biologically relevant interaction interface. We then determined by SEC-MALS that DdrC exists as a dimer in solution (Figure 4D , Table 4 ). Therefore, the interface identified by crystallography is most likely the dimerization interface.
SEC-MALS measurements of the oligomeric state of different DdrC variants in solution
Truncation . | Measured MW (kDa) . | Oligomeric state (n-mer) . |
---|---|---|
1–231 (FL) | 47.99 ± 4.14 | 1.90 ± 0.16 |
1–98 (NTD) | 19.67 ± 1.15 | 1.84 ± 0.11 |
99–231 (CTD) | 28.63 ± 5.06 | 1.96 ± 0.35 |
Truncation . | Measured MW (kDa) . | Oligomeric state (n-mer) . |
---|---|---|
1–231 (FL) | 47.99 ± 4.14 | 1.90 ± 0.16 |
1–98 (NTD) | 19.67 ± 1.15 | 1.84 ± 0.11 |
99–231 (CTD) | 28.63 ± 5.06 | 1.96 ± 0.35 |
From the crystal structure of DdrC, it appears that the protein is composed of two distinct domains, since the inter-and intramolecular contacts holding the dimer together are localized to either one of two regions: the N-terminal domain (NTD) that spans residues M1 to E110 and the C-terminal domain (CTD) that spans residues P111 to G231 (Figure 4A ). The first α-helix in the CTD contains a short stretch of 17 residues (P111-A126) that has an alternate conformation between the two DdrC chains. As such, we identified this section as a flexible ‘linker’. We generated the NTD and the CTD as independent proteins at these proposed domain boundaries (NTD: 1–110, CTD: 111–231), but the CTD exhibited very poor solubility. This poor solubility is likely the result of exposed hydrophobic residues, L117 and L121, that would otherwise interface with the NTD in the full-length protein. To improve solubility of the truncated domains, we generated NTD and CTD constructs with an altered domain boundary that neutralizes these exposed hydrophobic residues (NTD: 1–98, CTD: 99–231). These constructs were successfully expressed, purified, and remained stable in solution.
The presence of two distinct domains is supported by differential scanning fluorimetry (DSF) analysis, as full-length (FL) DdrC exhibits two melt peaks: one with a melting temperature ( T m ) of 40°C and one with a T m of 73°C (Figure 4B ). Each of the two proposed DdrC domains, NTD and CTD, were then analyzed under DSF separately as independent protein constructs. The thermal melt curves of the NTD and CTD constructs contained only one melt peak, each with a T m of 73°C and 37°C, respectively, matching the two T m values observed with FL DdrC. Both prominent melting events in the FL DdrC melt profile can be explained individually by the NTD and CTD melt profiles, demonstrating that the NTD and CTD are structurally distinct domains that fold independently of one another.
To investigate the possible role of each domain by structural homology, we queried the crystal structure of FL DdrC for structural similarity against the PDB databank. As expected, the NTD and CTD regions independently align to different structures, further supporting the hypothesis that the NTD and CTD of DdrC are distinct domains ( Supplementary Figure S7 ). The proteins that aligned most closely to the NTD were primarily members of the Dachshund homology domain (DHD) spanning residues ∼1–73 of DdrC. The CTD did not align to any specific domain family, but rather to an array of 3- and 4-helix bundles found broadly across the PDB. The degree of alignment of any characterized protein to either the NTD or CTD is too poor to extrapolate any meaningful functional significance.
To facilitate crystallization of FL DdrC, it was necessary to first optimize the buffer, pH and salt conditions of the protein storage solution for optimal protein stability ( Supplementary Figure S8 ). When storage conditions were sub-optimal, crystallization trials of WT FL DdrC did not yield any DdrC crystals, except on one non-reproducible occasion where a third crystal form was identified with a different unit cell and symmetry from the two previous crystal structures (PDB: 8U1J). Structure determination by MR revealed a DdrC dimer that was only comprised of residues 1–97 from the NTD (Figure 4C ). The remaining CTD residues (98–231) lacked any significant electron density. The relatively low refinement R -factors indicate that the NTD dimer model agrees with the experimental data and that the remaining 58% of the protein is either highly flexible under these conditions, or completely absent from the crystal lattice (Table 3 ). Given the expected structure of the CTD, residues 99–231 would not have fit into the observed crystal lattice ( Supplementary Figure S6 ), so it appeared as though DdrC was somehow crystallized without the CTD domain. The source crystallization drop was analyzed by SDS-PAGE, and it was confirmed that DdrC was indeed proteolytically digested in the drop, resulting in an apparent molecular weight that is consistent with residues 1–98 (Figure 4C ). This digestion was likely the result of an unknown contaminating protease. The high resistance of the NTD against proteolytic degradation is consistent with the previous observation that the NTD is more resistant to thermal denaturation than the CTD (Figure 4B ), implying that the NTD has high structural stability compared to the CTD.
The NTD crystal structure demonstrates that residues 1–98 of DdrC dimerize independently from the rest of the protein. In fact, the NTD dimer in the truncated crystal structure was nearly identical to the NTD dimer in the FL crystal structure (RMSD: 1.36Å). To validate that the NTD dimerizes in solution and not just in a stabilized crystal lattice, DdrC was expressed explicitly as a 1–98 truncation and purified. Upon measuring the oligomeric state by SEC-MALS, we found that the NTD does in fact dimerize in solution (Figure 4D , Table 4 ).
Despite the low stability of the CTD domain, we cannot rule out the possibility that the CTD can also dimerize independently of the rest of the protein. Attempts to express the CTD on its own (residues 99–231) yielded insoluble protein aggregates. However, we were able to express the CTD domain with a monomeric OCR (mOCR) fusion tag. The CTD remained stable and soluble following cleavage and removal of the fusion tag under optimized buffer conditions. SEC-MALS analysis revealed that, like the NTD, the CTD also dimerizes independently in solution (Figure 4D , Table 4 ).
Attempts to crystalize the CTD explicitly yielded no crystals, but it would be safe to assume that the dimerization interface of the CTD on its own would be the same interface as seen in the FL DdrC structure. Together, the NTD crystal structure with the SEC-MALS and DSF data demonstrate that the NTD and CTD domains fold and dimerize independently from each other.
When analyzed independently, the NTD and CTD each homodimerize symmetrically via a C2 axis of rotation. Curiously, the full DdrC homodimer comprising both NTD and CTD domains is itself asymmetric, with no global axes of symmetry between the two chains (Figure 5A ). In the context of the FL DdrC dimer, the C2 axes of the NTD and CTD are offset from each other by 46°. The same asymmetric structure is seen in both crystal forms despite different protein contacts and different chemical environments. So, this asymmetry is likely an endogenous structural feature, and not just a crystallographic artifact. Studying the nature of this asymmetry may reveal a molecular mechanism for DdrC function.
Analysis of structural asymmetry in the DdrC homodimer. ( A ) DdrC homodimer structure with highlighted midpoint positions between pairs of opposing homodimer atoms. Symmetry axes were fit by Principal Component Analysis (PCA) of midpoint positions for each domain. ( B ) Torsion angle differences (ΔΦ + ΔΨ) between both chains of the DdrC homodimer. The residues that most contribute to global asymmetry are indicated (★). ( C ) Loaded and relaxed conformations of the α6 helix from chains A and B, respectively. ( D ) Holding clasp residues in the disengaged and ( E ) engaged states. ( F ) Static forces in the DdrC homodimer counteract each other in a loaded mousetrap mechanism.
Comparing the internal coordinates between both DdrC chains, it is clear that the asymmetry can be attributed to only five residues within the interdomain region: residues 120–125 (Figure 5B ). The torsion angle differences between both DdrC chains at residues 120–125 account for all the global asymmetry in the DdrC dimer. In one DdrC chain, residues 120–125 lie within an intact helix, α6, which spans residues 111–136. In the other DdrC chain, α6 is deformed at residues 120–125, breaking the helix into two separate segments (Figure 5C ). Since the α6 residues clearly have a propensity to form a helix, we can assume that the broken α6 helix is under tension, like a bent spring.
The tension in the bent spring appears to be counteracted by a strong network of salt-bridges and H-bonds between the NTD and the CTD on only one face of the dimer (Figure 5E ). This ‘holding clasp’ mechanism involves the NTD from chain A (Gln-103, Glu-106) and the CTD from chain B (Arg-120, Asp-174). Meanwhile, the analogous clasp on the opposite face of the dimer is disengaged as the residues are not within range to form this interaction network (Figure 5D ). Interestingly, in the 7UDI crystal structure, the electron density is missing for residues 157–174 on the disengaged face of the dimer, suggesting that the CTD residues surrounding the holding clasp may be flexible when Asp-174 is not engaged.
We hypothesize that static forces within the DdrC homodimer counteract each other in a loaded mousetrap mechanism (Figure 5F ). A deformed α6 helix forms a loaded spring while an H-bond network forms a clasp that holds the spring under tension. It is possible that the mousetrap mechanism is a method for storing potential energy that can be used by DdrC to carry out its biological functions. If this is the case, then the mechanism must be triggered in response to a specific biochemical signal. This signal may be the terminal ends of DNA strands as we have shown that DdrC binds to ss- and ds-breaks with high affinity.
To investigate the structural basis of DNA nick detection by DdrC, it is necessary to determine the structure of DdrC in complex with DNA. Ongoing efforts to co-crystalize DdrC with different DNA ligands have yet to produce diffracting crystals; however, the apo-structure of DdrC hints at the location of two possible DNA binding sites. The electrostatic surface potential of the DdrC dimer reveals two large patches of partial positive charge on the surface (Figure 6A ). Even though both potential binding sites involve the same residues from opposing chains, they each have a different shape due to the asymmetry of the DdrC dimer. One possible binding site appears to be in an ‘open’ conformation, while the other is in a ‘closed’ conformation.
Prediction of DdrC-DNA complex formation. ( A ) Surface representation of the DdrC homodimer colored by domains (center) and by electrostatic surface potential (left and right). Two large surface patches of positive electrostatic potential are highlighted (°,•). The positively-charged residues corresponding to each patch are indicated on a DdrC domain map (bottom). ( B ) Computational model of DdrC bound to a DNA duplex with no internal lesions. Axes of symmetry are shown corresponding to the NTD (yellow) and the CTD (green). ( C ) Electrostatic surface of contacting residues in the CTD and ( D ) the NTD. ( E ) Conformation of the α6 helices in the DdrC-DNA complex. ( F ) Computational model of DdrC bound to two DNA duplexes with one single-strand break each. ( G ) Contacting residues in the CTD and ( H ) the NTD. ( I ) Conformation of the α6 helices in the dual-nick complex.
We used a rigid protein, flexible DNA docking algorithm to dock a single dsDNA duplex onto a DdrC dimer ( 37 ). The entire complex was then minimized in Rosetta to sample optimal protein–DNA contacts. As expected, the dsDNA duplex docked to one of the positive patches on DdrC. Of the two possible sites, the DNA docked to the ‘open’ binding site (Figure 6B ). In this binding mode, most of the protein–DNA interactions are mediated by the CTD (Figure 6C ). Residues within the flexible CTD clasp make non-sequence-specific contacts with the minor groove to the DNA duplex. The DdrC clasp is slightly deforming the duplex as it presses the DNA into the NTD, where some contacts between the NTD and the DNA backbone are made (Figure 6D ). In this conformation, the α6 helices of DdrC are still in the ‘loaded’ state, so any tension that may have been stored in the ‘loaded spring’ has not been released (Figure 6E ). Since all of the DdrC-DNA contacts in this model involve only internal DNA nucleotides, as opposed to ss- or ds-break sites, we hypothesize that this binding mode represents a state of lesion scanning. In this state, DdrC may be scanning for ss- or ds-breaks by deforming the DNA and interrogating the DNA duplex for a specific mechanical response. To fully understand the mechanism of nick detection, a structure of DdrC bound to nicked DNA is also required.
Accurate prediction of a nicked dsDNA complex is challenging to accomplish using traditional docking methods because the addition of a single nick greatly increases the structural degrees of freedom of the DNA duplex. The docking algorithm would need to cover a very large conformational space for both the DdrC dimer and the nicked duplex in order to find the correct docking pose. Recent advances in structure prediction now allow for de novo prediction of protein-nucleic acid complexes from sequence using a trained neural-net model ( 52 ). One such algorithm, RF2NA, was used to predict the structure of a DdrC dimer in complex with a nicked DNA duplex, which was then minimized using Rosetta Relax (Figure 6F ). Unlike the unbroken DNA duplex, the nicked DNA was docked to the ‘closed’ pocket of the DdrC dimer. An unbroken DNA duplex would be unable to fit in the closed pocket, but the introduction of a single nick appears to increase DNA flexibility in such a way that binding to the closed pocket becomes possible. When a nick is present, the DNA duplex lends itself to significant deformation by residues within the CTD clasp. This deformation is primarily mediated by Lys-170 on DdrC as it protrudes into the DNA duplex and disrupts base-pair contacts, while forming a π-cation interaction with the face of a DNA base (Figure 6G ). The duplex deformation by the CTD also allows for nick-specific contacts to occur within the NTD. Most notably, Arg-14 is predicted to form a π-cation interaction with the terminal DNA base on the 5′ end of the nick, while Arg-81 forms a positively charged binding pocket for DNA backbone atoms on the 3′ end of the nick, including the terminal 3′OH itself (Figure 6H ). It is of note that the terminal Phosphate group on the 5′ end is not predicted to form any polar contacts with DdrC, and so it does not appear to be important for nick detection. This feature of the model matches experimental observations, as nick-mediated DNA compaction by DdrC is unimpaired when all 5′ Phosphates are removed (Figure 2D ).
Another interesting feature of the DdrC-nicked DNA structure is the change in global symmetry. RF2NA predicts a symmetric binding conformation of DdrC, where both the binding sites are in a ‘closed’ conformation, allowing for binding to two identical nicked duplexes. In this conformation, both α6 helices in the DdrC dimer are predicted to be in the relaxed state, suggesting that the tension in α6 has been released (Figure 6I ). In order for both DdrC binding sites to be in the closed states, it is necessary for the CTD dimer interface to be disrupted. This is the prediction made by the RF2NA algorithm. Given the low number of intermolecular contacts holding the CTD dimer together, and given the low thermal stability of the CTD alone (Figure 4B ), it is clear that the CTD dimer contacts are very weak. It is therefore plausible that the CTD interface becomes broken during the nick detection process, leading to two symmetric DNA binding sites.
Together, the unbroken DNA and nicked DNA binding models hint at a possible nick detection mechanism (Figure 7A ). First, DdrC binds to unbroken DNA along the open face of the dimer, ‘scanning’ for a nick. DdrC scans for nicks by attempting to deform the duplex using energy stored in the loaded α6 helix. If a ss-break is present, the DNA will lend itself to deformation and the ‘open face’ of DdrC will adopt a closed conformation. This conformational change causes the opening of a second DNA binding site on the opposite face of the dimer that was previously closed. The newly opened binding site is now free to scan for a second nick. Once a second nick is detected, the binding site closes, trapping two DNA nicks in a conformation that is symmetric about a C2 axis of symmetry. In this conformation, the DNA duplexes are placed in a perpendicular orientation to each other, which topologically mimics a supercoiling writhe point. This duplex crossover structure would mimic a positive supercoil if both nicks are on the same DNA strand, and would mimic a negative supercoil if the nicks are on opposite strands ( Supplementary Figure S9 ). By introducing a topological writhe point for every pair of ss-breaks, DdrC progressively compacts circular DNA to a degree that is proportional with the amount of DNA damage. In other words, more ss-breaks lead to more compaction (Figure 7B ).
Proposed mechanism of DNA lesion detection and DNA topology modulation by DdrC. ( A ) Mechanism of dual nick detection. ( B ) Mechanism of ss-break-mediated DNA compaction. ( C ) Proposed binding mode of DdrC to supercoiled DNA. ( D ) Mechanism of ds-break-mediated DNA circularization. High-affinity binding events requiring a conformational change in DdrC are labeled (★).
In addition to compacting nicked DNA via supercoil-like structures, we have observed that DdrC binds directly to DNA that is already in a supercoiled state prior to binding. The band shift that occurs when DdrC interacts with supercoiled DNA is progressive, meaning that the band becomes more shifted with increasing concentrations of DdrC up to a saturation point ( Supplementary Figure S1 ). This behavior implies that there are many possible DdrC binding sites on the supercoiled plasmid that become progressively saturated. Since the affinity of the supercoiled binding event is similar to that of nicked DNA, the binding mechanisms are probably also similar. In nicked DNA, DdrC recognizes lesions through duplex deformations. We hypothesize that DdrC recognizes regions of supercoiled DNA where similar duplex deformations arise spontaneously ( Supplementary Figure S10 ). It is well understood that supercoiled DNA adopts local DNA deformations such as wrinkles, bubbles, kinks and slips to alleviate the stresses of over- or under-winding ( 50 , 52–55 ). In many of these deformations, base pair and base-stacking contacts in the DNA are disrupted, exposing unpaired bases to solvent. Since DdrC recognizes DNA nicks through DNA deformations and π interactions with exposed bases, it is possible that DdrC recognizes some specific local deformation in supercoiled DNA via the same mechanism (Figure 7C ).
Finally, we have demonstrated that DdrC circularizes DNA via direct binding to ds-breaks. A plausible structure of the ds-break complex can be generated from the nicked complex by truncating the DNA to a blunt-end at the site of the nick and minimizing the resulting model ( Supplementary Figure S11A, B ). If we follow the proposed steps of nick detection with ds-breaks instead of ss-breaks, we inevitably arrive at a mechanism for DNA circularization by DdrC (Figure 7D ). Like with the nicked DNA substrate, the CTD forms sequence-independent contacts with the duplex and pushes the ds-break into the NTD of DdrC. The NTD then makes end-specific contacts with the 5′ and 3′ terminal ends of the DNA. According to this model, a 5′ overhang would disrupt the ability of the NTD to properly engage the 3′ end of the duplex ( Supplementary Figure S11C ). So, we would expect that a 5′ overhang should reduce the affinity of DdrC to ds-breaks. This is precisely what we have observed experimentally (Figure 3A ).
In our gel-based DdrC-DNA binding assays, we can resolve two distinct binding events: one that is dependent on, and one that is independent from the presence of DNA lesions; referred to as the high-affinity (HA) and low affinity (LA) binding events, respectively ( Supplementary Figure S14 ). The HA binding event occurs at relatively low DdrC concentrations (∼10–100 nM) and requires the presence of two or more dsDNA lesions in the form of single-strand or double-strand breaks. The HA mode of binding results in a complex with a fixed stoichiometric ratio of DdrC to DNA as evidenced by a band shift to a discrete position on the gel. This DdrC-DNA species likely corresponds to the DdrC-DNA complex in Figure 6F , as this mode of binding localizes DdrC to a specific site on the DNA. On the other hand, the LA binding event, which requires higher concentrations of DdrC (>1 μM), is indicative of non-specific interactions and occurs with any form of dsDNA, irrespective of the presence of lesions or DNA topology. This binding mode leads to a substantial upward gel shift of the DNA, often preceded by a ‘smearing’ of the band. The smearing suggests that the LA mode of binding involves a variable stoichiometric ratio of DdrC to DNA, potentially involving multiple DdrC molecules per plasmid. We hypothesize that the ‘scanning’ DdrC-DNA complex shown in Figure 6B corresponds to the LA binding event on the gel, as it allows DdrC to bind at any position on the plasmid, not just at the site of damage.
By monitoring the formation of the HA and the LA species, it is possible to independently assay the lesion-specific and non-specific binding activities of DdrC. If our computational models are correct, we should be able to disrupt these two binding activities selectively through targeted mutagenesis. To disrupt non-lesion-specific interactions (corresponding to the LA species), we targeted highly conserved residues on DdrC that are predicted to directly contact DNA in both our intact and nicked DNA models. Four residues (Arg-128, Arg-142, Arg-164 and Lys-170) were identified as potential candidates for disruption, as they are predicted to be involved in ionic interactions with the DNA backbone in both scanning and nick-binding states (Figure 8A ). These interactions were disrupted by substituting all four residues with alanine (CTD-mut) or by deleting the entire C-terminal domain (ΔCTD) (Figure 8B ).
Targeted disruption of DdrC-DNA interactions. ( A ) Predicted protein–DNA interface of the HA and LA complexes at the NTD and CTD domains. Residues that were targeted for mutagenesis are highlighted and labeled. ( B ) Summary of DdrC variants harboring disruptive amino acid substitutions and deletions. ( C ) DNA binding activity of each DdrC variant measured against pUC19 plasmid in 3 different topological states: nicked, linear and supercoiled. For each DdrC/plasmid combination, 2 nM of plasmid was incubated with varying concentrations of DdrC. The relative fractions of DNA were measured as bound in either the HA, LA or unbound states. The total percentage of bound DNA is plotted here as the sum of HA and LA fractions. ( D ) Survival of different D.radiodurans R1 strains in response to varying UV-C doses. The surviving fraction was measured in triplicate as CFU counts relative to unirradiated bacteria. All D.radiodurans strains harbor a ΔuvsE genetic background as well as a Deinococcus expression plasmid with the corresponding ddrC ORF under constitutive expression from the PDR_1261 promoter.
To disrupt lesion-specific interactions (corresponding to the HA species), we targeted residues predicted to selectively interact with nicked DNA, while not interacting with un-nicked DNA. Two such residues, Arg-14 and Arg-81, were chosen. Arg-14 is predicted to interact with the face of the terminal base on the 5′ end of the nick through a π-cation interaction, while Arg-81 is part of a positively charged binding pocket for DNA backbone atoms on the 3′ end of the nick, including the terminal 3′OH itself (Figure 8A ). To disrupt these interactions, we either substituted the residues with alanine (NTD-mut) or deleted the entire N-terminal domain (ΔNTD) (Figure 8B ).
The DNA binding characteristics of the DdrC variants were assayed against both linear, nicked, and supercoiled pUC19 (Figure 8C ). Disrupting the non-specific interactions (CTD-mut and ΔCTD) led to a disruption of all DNA binding activity, both LA and HA. The binding appears to be disrupted equally across all three topological forms of pUC19 plasmid. This finding is consistent with our computational models that predict that the CTD domain harbors the ‘core’ DNA binding residues of DdrC. These residues are required for binding to any DNA ligand.
Disrupting the lesion-specific interactions by removal of the NTD domain (ΔNTD) resulted in the complete loss of the HA binding event, indicating that the NTD domain is responsible for lesion recognition. Interestingly, there was still detectable LA binding activity with the ΔNTD protein. This demonstrates that the CTD alone is sufficient for DNA binding but not for lesion detection.
Disruption of the two predicted nick-recognition residues in NTD domain (NTD-mut) resulted in the complete loss of the HA binding event in the case of nicked pUC19. Interestingly, the mutations had a milder effect on HA binding in the case of linear and supercoiled pUC19. This finding suggests that residues R14 and/or R81 specifically bind to ss-breaks over ds-breaks. This is supported by the computational model of DdrC bound to ds-breaks, where only Arg-14 participates in recognition of the ds-break, but not Arg-81 ( Supplementary Figure S11B ).
Although the lesion recognition residues are located in the NTD, the NTD alone cannot effectively stabilize the DdrC-DNA complex, as evidenced by the complete lack of HA or LA binding in the ΔCTD variant. The mechanism of lesion detection by DdrC likely requires proper positioning of the DNA duplex by the CTD for the lesion to make proper contact with the NTD.
To verify whether DNA binding and lesion recognition are important features of DdrC function in vitro , we measured the effect of disruptive ddrC mutations on UV-C resistance in live Deinococcus bacteria. In D. radiodurans R1 , a deletion of the ddrC gene only exhibits a modest 10-fold reduction in UV-C resistance ( 25 ). However, when combined with a deletion of the uvsE gene, a ddrC deletion is 50–100-fold more sensitive to UV-C radiation compared to a knockout of uvsE alone ( 25 ). To maximize the signal-to-noise ratio of measurements quantifying the phenotype of ddrC point mutations, all functional assays were conducted in D. radiodurans with a ΔuvsE genetic background. The functional status of different ddrC viariants were measured by expressing ddrC from a plasmid in ΔuvsEΔddrC D. radiodurans then assaying for its ability to restore UV-C resistance relative to a ΔuvsE baseline phenotype.
As expected, D. radiodurans ΔuvsE harboring an empty expression vector becomes more sensitive to UV-C when a ddrC deletion is introduced (Figure 8D , Table 5 , Supplementary Figure S17 ). Under our experimental conditions, we measure a 94% loss in UV-resistance at doses of 200 J/m 2 or higher. This loss in UV resistance can then be restored by complimentary expression of WT ddrC . In fact, the ddrC complement appears to be 210% more UV resistant than the ΔuvsE baseline at 200 J/m 2 . When the same complementation experiment is performed using mutants of ddrC , neither the NTD-mut or the CTD-mut variants of ddrC can restore UV resistance to a ddrC knockout strain, indicating that these mutants disrupt ddrC function. When compared to WT ddrC , the NTD- and CTD-mut variants exhibit a 99.8% and 99.7% loss of UV resistance respectively at a UV dose of 200 J/m 2 . In vitro, we observe that the NTD-mut lacks the ability to bind ss-breaks while the CTD-mut is deficient in DNA binding overall. From these observations, we can conclude that both DNA binding and lesion recognition are essential features of ddrC function.
Surviving fraction of different D.radiodurans R1 strains in response to varying UV-C doses
(J/m ) . | + empty vector . | + empty vector . | + ddrC (WT) . | + ddrC (NTD-mut) . | + ddrC (CTD-mut) . |
---|---|---|---|---|---|
0 | 1.00 ± 0.09 × 10 | 1.00 ± 0.13 × 10 | 1.00 ± 0.06 × 10 | 1.00 ± 0.12 × 10 | 1.00 ± 0.26 × 10 |
28.7 | 5.19 ± 0.48 × 10 | 7.52 ± 1.35 × 10 | 6.07 ± 1.12 × 10 | 2.68 ± 0.44 × 10 | 3.18 ± 0.60 × 10 |
57.4 | 3.09 ± 0.70 × 10 | 5.09 ± 1.02 × 10 | 4.39 ± 1.22 × 10 | 6.89 ± 0.85 × 10 | 9.39 ± 2.51 × 10 |
86.1 | 1.04 ± 0.17 × 10 | 2.68 ± 0.46 × 10 | 2.11 ± 0.39 × 10 | 6.36 ± 3.47 × 10 | 8.27 ± 2.36 × 10 |
114.8 | 6.63 ± 1.08 × 10 | 9.22 ± 0.97 × 10 | 2.10 ± 0.38 × 10 | 2.02 ± 1.42 × 10 | 1.33 ± 0.71 × 10 |
143.5 | 9.37 ± 5.01 × 10 | 6.26 ± 1.51 × 10 | 5.33 ± 2.48 × 10 | 6.19 ± 5.59 × 10 | 3.19 ± 2.06 × 10 |
172.2 | 3.77 ± 1.00 × 10 | 7.26 ± 3.27 × 10 | 1.92 ± 0.24 × 10 | 8.16 ± 3.52 × 10 | 5.25 ± 2.10 × 10 |
200.9 | 1.71 ± 0.32 × 10 | 1.07 ± 0.75 × 10 | 5.37 ± 0.86 × 10 | 1.13 ± 0.53 × 10 | 1.81 ± 0.27 × 10 |
(J/m ) . | + empty vector . | + empty vector . | + ddrC (WT) . | + ddrC (NTD-mut) . | + ddrC (CTD-mut) . |
---|---|---|---|---|---|
0 | 1.00 ± 0.09 × 10 | 1.00 ± 0.13 × 10 | 1.00 ± 0.06 × 10 | 1.00 ± 0.12 × 10 | 1.00 ± 0.26 × 10 |
28.7 | 5.19 ± 0.48 × 10 | 7.52 ± 1.35 × 10 | 6.07 ± 1.12 × 10 | 2.68 ± 0.44 × 10 | 3.18 ± 0.60 × 10 |
57.4 | 3.09 ± 0.70 × 10 | 5.09 ± 1.02 × 10 | 4.39 ± 1.22 × 10 | 6.89 ± 0.85 × 10 | 9.39 ± 2.51 × 10 |
86.1 | 1.04 ± 0.17 × 10 | 2.68 ± 0.46 × 10 | 2.11 ± 0.39 × 10 | 6.36 ± 3.47 × 10 | 8.27 ± 2.36 × 10 |
114.8 | 6.63 ± 1.08 × 10 | 9.22 ± 0.97 × 10 | 2.10 ± 0.38 × 10 | 2.02 ± 1.42 × 10 | 1.33 ± 0.71 × 10 |
143.5 | 9.37 ± 5.01 × 10 | 6.26 ± 1.51 × 10 | 5.33 ± 2.48 × 10 | 6.19 ± 5.59 × 10 | 3.19 ± 2.06 × 10 |
172.2 | 3.77 ± 1.00 × 10 | 7.26 ± 3.27 × 10 | 1.92 ± 0.24 × 10 | 8.16 ± 3.52 × 10 | 5.25 ± 2.10 × 10 |
200.9 | 1.71 ± 0.32 × 10 | 1.07 ± 0.75 × 10 | 5.37 ± 0.86 × 10 | 1.13 ± 0.53 × 10 | 1.81 ± 0.27 × 10 |
Interestingly, both the NTD-mut and CTD-mut strains exhibit significantly lower UV resistance compared to the ddrC knockout strain, demonstrating that the presence of DdrC protein with a DNA-binding deficiency is more detrimental to UV resistance than a deletion of the ddrC gene alone (Figure 8D , Table 5 , Supplementary Figure S17 ). When compared to an empty vector, the plasmids harboring the NTD and CTD-mutants confer in an 89% and 83% loss in UV resistance respectively at a UV dose of 200 J/m 2 . This dominant-negative phenotype suggests that DdrC may have additional functions beyond DNA binding and lesion recognition. One possible explanation for the dominant-negative phenotype is that DdrC may play a role in recruiting other factors to DNA via protein–protein interactions. Consequently, a semi-functional DdrC mutant deficient in DNA binding but proficient in protein binding might interfere with the function of other DNA repair proteins by sequestering them away from DNA.
In this study, we demonstrate that DdrC binds to duplex DNA at sites of ss- and ds-breaks. The observed binding characteristics strongly suggest that each functional unit of DdrC contains two DNA binding sites.
This finding is corroborated by our crystal structure of DdrC, as there are two large patches of positive electrostatic surface potential on the DdrC dimer, which hint at the location of two DNA binding pockets. Interestingly, the two potential binding sites are structurally different, despite involving the same residues from opposing DdrC chains. This asymmetry appears to be an evolved feature of the DdrC dimer, as it has been observed by us and others under different crystal lattices ( 26 ).
Our computational modeling experiments suggest that in this asymmetric state, one binding site is in an open conformation while the other is in a closed conformation. We propose that DdrC transiently binds to DNA duplex at arbitrary positions via the open binding site and scans the duplex for lesions. These interactions primarily involve DdrC residues from the CTD domain.
The presence of a ss- or ds-break allows the DNA to form favorable interactions with DdrC residues in the NTD, triggering a conformational change in DdrC. This change in conformation opens the second binding site on DdrC, through which the DdrC dimer can scan for and trap a second DNA break. This mechanism is supported by functional assays with DdrC variants, which harbor disruptive amino acid substitutions and deletions. The findings indicate that the CTD involves the core DNA binding residues while the NTD contains residues that interact specifically with DNA lesions. We showed that mutation of these residues disrupts UV-C resistance in vivo , thus demonstrating that both DNA binding and lesion recognition are necessary features of DdrC function.
It is possible that the conformation change required for nick detection is driven by stored tension forces in the DdrC dimer. In the asymmetric scanning state, the α6 helix connecting the NTD to the CTD is significantly deformed in one DdrC monomer, but not in the other. The straightening of this helix would result in the re-symmetrisation of the DdrC dimer. We also observed a network of ionic interactions between the NTD of one DdrC chain and the CTD of the other. This ‘clasping’ mechanism appears to be holding the DdrC dimer in an asymmetric state. The presence of such a robust salt-bridge network may be taken as evidence that the clasp mechanism evolved to counteract a strong force in the opposite direction. As such, we propose that there is potential energy stored in the deformation of the α6 helix. The presence of a ss-break may trigger the release this energy, causing a conformation change in DdrC and the trapping of a DNA nick.
This is similar to the mechanism of nick detection in human cells, where single-strand breaks are rapidly bound by the signalling factor PARP-1 ( 56 ). When bound internally on an un-nicked DNA ligand, PARP-1 exists as a loosely associated string of protein domains with high potential energy. The protein interrogates the DNA for ss-breaks through dimerization interactions between the F1and F2 domains. Only in the presence of a ss-break can the F2 domain twist the DNA in such a way to enable a stabilizing π-stacking interaction between a Phe residue on PARP-1 and the face of a nucleotide base that is exposed at the site of the nick (PDB 2N8A ( 57 )). The proper dimerization of F1 and F2 initiates a ‘structure collapse’ where the other PARP-1 domains associate with F1, F2 and the DNA surface to collapse into a low-energy state. The structure collapse results in a very high affinity interaction with the ss-break. The fully formed PARP-1 assembly then recruits DNA repair factors to the site of damage and triggers the DNA-damage response cascade.
The proposed mechanism for nick detection by DdrC is also similar to the mechanism of inter-base adduct recognition by the Rad4/XPC complex ( 58 , 59 ). Like DdrC, Rad4/XPC binds DNA in a scanning conformation and interrogates the DNA duplex for lesions by attempting to adopt a lower-energy protein–DNA conformation. In the case of Rad4/XPC, the complex scans by flipping out DNA bases. Although base flipping is possible for both damaged and undamaged DNA, the kinetic barrier of this operation is much lower for DNA at the site of an inter-base adduct like a thymine dimer.
Unlike PARP-1 and Rad4/XPC, DdrC senses and traps two DNA lesions for each structural unit of DdrC. The consequence of trapping pairs of DNA breaks effectively results in the circularization of linear DNA and the compaction of nicked DNA. Furthermore, the degree of DNA compaction increases with the degree of DNA damage. We have shown this to be the case using a series of precisely nicked and linearized plasmids. It makes sense why this observed behavior of DdrC would be useful to Deinococcus bacteria under DNA damaging conditions. Two ss-breaks on opposite strands could become ds-breaks if they are close enough in proximity. The immobilization of ss-breaks and compaction of nicked DNA by DdrC could prevent ss-breaks from becoming ds-breaks as ss-breaks accumulate across the genome. In the event of a ds-break, DdrC is able to bridge the DNA ends to prevent end diffusion. In addition, the trapping of DNA lesions allows Deinococcus to control the supercoiling state of its genome, even in the presence of ss- and ds-breaks.
The plasmid compaction behavior of DdrC has been observed previously, although the dependence of this process on ss-breaks has not yet been reported. Using TEM and AFM techniques, DdrC has been shown to directly promote plasmid compaction and circularization in vitro ( 25 , 26 ). In vivo , DdrC has been observed to co-localize with the compact nucleoid DNA structures in D. radiodurans following rapid expression in response to γ-radiation ( 25 ). These observations led to the proposal that DdrC is a novel nucleoid-associated protein (NAP) that maintains a compact nucleoid structure following extreme DNA damage ( 25 , 26 ). Our results support this claim. We propose that DdrC behaves as a lesion-specific NAP that neutralizes ss-breaks and compacts the nucleoid to a degree that is proportional with the amount of DNA damage present. This would aid other NAPs in maintaining a compact nucleoid even in the presence of DNA damage. Furthermore, we observed a dominant-negative behaviour of DdrC mutations that are deficient in DNA binding and nick-recognition, suggesting that DdrC may have other functions beyond DNA binding. It is possible that through protein-protein interactions, DdrC may be interacting with other NAPs or may be recruiting other repair factors to the sites of DNA damage.
Refined crystal structures and structure factors were deposited to the Protein Data Bank under the accessions 7UDI (DOI: 10.2210/pdb7UDI/pdb), 8U0G (DOI: 10.2210/pdb8U0G/pdb) and 8U1J (DOI: 10.2210/pdb8U1J/pdb). Raw diffraction images for the crystal structures 7UDI, 8U0G and 8U1J were deposited to Zenodo under the accession numbers 10022358, 8302395 and 8309780 respectively (DOI: 10.5281/zenodo.10022358, 10.5281/zenodo.8302395, 10.5281/zenodo.8309780). Computational structure models were deposited to ModelArchive under the accession numbers ma-nmyn0 (DOI: 10.5452/ma-nmyn0), ma-urph3 (DOI: 10.5452/ma-urph3), ma-50nj9 (DOI: 10.5452/ma-50nj9) and ma-otnza (DOI: 10.5452/ma-otnza).
Supplementary Data are available at NAR Online.
Part of the research described in this paper was performed using beamlines CMCF-ID and CMCF-BM at the Canadian Light Source, a national research facility of the University of Saskatchewan, which is supported by the Canada Foundation for Innovation (CFI), the Natural Sciences and Engineering Research Council (NSERC), the National Research Council (NRC), the Canadian Institutes of Health Research (CIHR), the Government of Saskatchewan, and the University of Saskatchewan.
Natural Sciences and Engineering Research Council of Canada [2008R00075]. Funding for open access charge: Natural Sciences and Engineering Research Council of Canada [2008R00075].
Conflict of interest statement . None declared.
Slade D. , Radman M. Oxidative stress resistance in Deinococcus radiodurans . Microbiol. Mol. Biol. Rev. 2011 ; 75 : 133 – 191 .
Google Scholar
Cox M.M. , Battista J.R. Deinococcus radiodurans - the consummate survivor . Nat. Rev. Microbiol. 2005 ; 3 : 882 – 892 .
Gaidamakova E.K. , Sharma A. , Matrosova V.Y. , Grichenko O. , Volpe R.P. , Tkavc R. , Conze I.H. , Klimenkova P. , Balygina I. , Horne W.H. et al. . Small-molecule Mn antioxidants in Caenorhabditis elegans and deinococcus radiodurans supplant MnSOD enzymes during aging and irradiation . mBio . 2022 ; 13 : e0339421 .
Daly M.J. , Gaidamakova E.K. , Matrosova V.Y. , Kiang J.G. , Fukumoto R. , Lee D.-Y. , Wehr N.B. , Viteri G.A. , Berlett B.S. , Levine R.L. Small-molecule antioxidant proteome-shields in Deinococcus radiodurans . PLoS One . 2010 ; 5 : e12570 .
Sharma A. , Gaidamakova E.K. , Grichenko O. , Matrosova V.Y. , Hoeke V. , Klimenkova P. , Conze I.H. , Volpe R.P. , Tkavc R. , Gostinčar C. et al. . Across the tree of life, radiation resistance is governed by antioxidant Mn2+, gauged by paramagnetic resonance . Proc. Natl. Acad. Sci. U.S.A. 2017 ; 114 : E9253 – E9260 .
Zahradka K. , Slade D. , Bailone A. , Sommer S. , Averbeck D. , Petranovic M. , Lindner A.B. , Radman M. Reassembly of shattered chromosomes in Deinococcus radiodurans . Nature . 2006 ; 443 : 569 – 573 .
Slade D. , Lindner A.B. , Paul G. , Radman M. Recombination and replication in DNA repair of heavily irradiated Deinococcus radiodurans . Cell . 2009 ; 136 : 1044 – 1055 .
Magerand R. , Rey P. , Blanchard L. , de Groot A. Redox signaling through zinc activates the radiation response in Deinococcus bacteria . Sci. Rep. 2021 ; 11 : 4528 .
Narasimha A. , Basu B. New insights into the activation of radiation desiccation response regulon in Deinococcus radiodurans . J. Biosci. 2021 ; 46 : 10 .
Blanchard L. , Guérin P. , Roche D. , Cruveiller S. , Pignol D. , Vallenet D. , Armengaud J. , de Groot A. Conservation and diversity of the IrrE/DdrO-controlled radiation response in radiation-resistant Deinococcus bacteria . Microbiologyopen . 2017 ; 6 : e00477 .
Eugénie N. , Zivanovic Y. , Lelandais G. , Coste G. , Bouthier de la Tour C. , Bentchikou E. , Servant P. , Confalonieri F. Characterization of the radiation desiccation response regulon of the radioresistant bacterium Deinococcus radiodurans by Integrative Genomic analyses . Cells . 2021 ; 10 : 2536 .
Tanaka M. , Earl A.M. , Howell H.A. , Park M.-J. , Eisen J.A. , Peterson S.N. , Battista J.R. Analysis of Deinococcus radiodurans 's transcriptional response to ionizing radiation and desiccation reveals novel proteins that contribute to extreme radioresistance . Genetics . 2004 ; 168 : 21 – 33 .
Hua Y. , Narumi I. , Gao G. , Tian B. , Satoh K. , Kitayama S. , Shen B. PprI: a general switch responsible for extreme radioresistance of Deinococcus radiodurans . Biochem. Biophys. Res. Commun. 2003 ; 306 : 354 – 360 .
Narumi I. , Satoh K. , Cui S. , Funayama T. , Kitayama S. , Watanabe H. PprA: a novel protein from Deinococcus radiodurans that stimulates DNA ligation . Mol. Microbiol. 2004 ; 54 : 278 – 285 .
Devigne A. , Meyer L. , de la Tour C.B. , Eugénie N. , Sommer S. , Servant P. The absence of the RecN protein suppresses the cellular defects of Deinococcus radiodurans irradiated cells devoid of the PprA protein by limiting recombinational repair of DNA lesions . DNA Repair (Amst.) . 2019 ; 73 : 144 – 154 .
Devigne A. , Guérin P. , Lisboa J. , Quevillon-Cheruel S. , Armengaud J. , Sommer S. , Bouthier de la Tour C. , Servant P. PprA protein is involved in chromosome segregation via its physical and functional interaction with DNA gyrase in irradiated deinococcus radiodurans bacteria . mSphere . 2016 ; 1 : e00036-15 .
Kota S. , Charaka V.K. , Ringgaard S. , Waldor M.K. , Misra H.S. PprA contributes to deinococcus radiodurans resistance to nalidixic acid, genome maintenance after DNA damage and interacts with deinococcal topoisomerases . PLoS One . 2014 ; 9 : e85288 .
Iyer L.M. , Koonin E.V. , Aravind L Classification and evolutionary history of the single-strand annealing proteins, RecT, Redbeta, ERF and RAD52 . Bmc Genomics [Electronic Resource] . 2002 ; 3 : 8 .
Gutsche I. , Vujičić-Žagar A. , Siebert X. , Servant P. , Vannier F. , Castaing B. , Gallet B. , Heulin T. , de Groot A. , Sommer S. et al. . Complex oligomeric structure of a truncated form of DdrA: a protein required for the extreme radiotolerance of Deinococcus . Biochim. Biophys. Acta - Proteins Proteomics . 2008 ; 1784 : 1050 – 1058 .
Harris D.R. , Ngo K.V. , Cox M.M. The stable, functional core of DdrA from Deinococcus radiodurans R1 does not restore radioresistance in vivo . J. Bacteriol. 2008 ; 190 : 6475 – 6482 .
Sugiman-Marangos S.N. , Weiss Y.M. , Junop M.S. Mechanism for accurate, protein-assisted DNA annealing by Deinococcus radiodurans DdrB . Proc. Natl. Acad. Sci. U.S.A. 2016 ; 113 : 4308 – 4313 .
Bouthier de la Tour C. , Boisnard S. , Norais C. , Toueille M. , Bentchikou E. , Vannier F. , Cox M.M. , Sommer S. , Servant P. The deinococcal DdrB protein is involved in an early step of DNA double strand break repair and in plasmid transformation through its single-strand annealing activity . DNA Repair (Amst.) . 2011 ; 10 : 1223 – 1231 .
Xu G. , Lu H. , Wang L. , Chen H. , Xu Z. , Hu Y. , Tian B. , Hua Y. DdrB stimulates single-stranded DNA annealing and facilitates RecA-independent DNA repair in Deinococcus radiodurans . DNA Repair (Amst.) . 2010 ; 9 : 805 – 812 .
de la Tour C.B. , Mathieu M. , Servant P. , Coste G. , Norais C. , Confalonieri F. Characterization of the DdrD protein from the extremely radioresistant bacterium Deinococcus radiodurans . Extremophiles . 2021 ; 25 : 343 – 355 .
Bouthier de la Tour C. , Mathieu M. , Meyer L. , Dupaigne P. , Passot F. , Servant P. , Sommer S. , Le Cam E. , Confalonieri F. In vivo and in vitro characterization of DdrC, a DNA damage response protein in Deinococcus radiodurans bacterium . PLoS One . 2017 ; 12 : e0177751 .
Banneville A.-S. , Bouthier de la Tour C. , De Bonis S. , Hognon C. , Colletier J.-P. , Teulon J.-M. , Le Roy A. , Pellequer J.-L. , Monari A. , Dehez F. et al. . Structural and functional characterization of DdrC, a novel DNA damage-induced nucleoid associated protein involved in DNA compaction . Nucleic Acids Res. 2022 ; 50 : 7680 – 7696 .
Buzon B. , Grainger R. , Huang S. , Rzadki C. , Junop M.S. Structure-specific endonuclease activity of SNM1A enables processing of a DNA interstrand crosslink . Nucleic Acids Res. 2018 ; 46 : 9057 – 9066 .
Virtanen P. , Gommers R. , Oliphant T.E. , Haberland M. , Reddy T. , Cournapeau D. , Burovski E. , Peterson P. , Weckesser W. , Bright J. et al. . SciPy 1.0: fundamental algorithms for scientific computing in Python . Nat. Methods . 2020 ; 17 : 261 – 272 .
Vonrhein C. , Flensburg C. , Keller P. , Sharff A. , Smart O. , Paciorek W. , Womack T. , Bricogne G. Data processing and analysis with the autoPROC toolbox . Acta Crystallogr. Sect. D Biol. Crystallogr. 2011 ; 67 : 293 – 302 .
Liebschner D. , Afonine P.V. , Baker M.L. , Bunkóczi G. , Chen V.B. , Croll T.I. , Hintze B. , Hung L.-W. , Jain S. , McCoy A.J. et al. . Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix . Acta Crystallogr. Sect. D Struct. Biol. 2019 ; 75 : 861 – 877 .
Cowtan K. Completion of autobuilt protein models using a database of protein fragments . Acta Crystallogr. Sect. D Biol. Crystallogr. 2012 ; 68 : 328 – 335 .
Emsley P. , Lohkamp B. , Scott W.G. , Cowtan K. Features and development of Coot . Acta Crystallogr. Sect. D Biol. Crystallogr. 2010 ; 66 : 486 – 501 .
Huang P.-S. , Ban Y.-E.A. , Richter F. , Andre I. , Vernon R. , Schief W.R. , Baker D. RosettaRemodel: a generalized framework for flexible backbone protein design . PLoS One . 2011 ; 6 : e24109 .
Alford R.F. , Leaver-Fay A. , Jeliazkov J.R. , O’Meara M.J. , DiMaio F.P. , Park H. , Shapovalov M.V. , Renfrew P.D. , Mulligan V.K. , Kappel K. et al. . The Rosetta all-atom energy function for macromolecular modeling and design . J. Chem. Theory Comput. 2017 ; 13 : 3031 – 3048 .
Holm L. Dali server: structural unification of protein families . Nucleic Acids Res. 2022 ; 50 : W210 – W215 .
Jurrus E. , Engel D. , Star K. , Monson K. , Brandi J. , Felberg L.E. , Brookes D.H. , Wilson L. , Chen J. , Liles K. et al. . Improvements to the APBS biomolecular solvation software suite . Protein Sci. 2018 ; 27 : 112 – 128 .
Banitt I. , Wolfson H.J. ParaDock: a flexible non-specific DNA–rigid protein docking algorithm . Nucleic Acids Res. 2011 ; 39 : e135 – e135 .
Conway P. , Tyka M.D. , DiMaio F. , Konerding D.E. , Baker D. Relaxation of backbone bond geometry improves protein energy landscape modeling . Protein Sci. 2014 ; 23 : 47 – 55 .
Baek M. , McHugh R. , Anishchenko I. , Baker D. , DiMaio F. Accurate prediction of nucleic acid and protein-nucleic acid complexes using RoseTTAFoldNA . Nat. Methods . 2024 ; 21 : 117 – 121 .
Chen A. , Sherman M.W. , Chu C. , Gonzalez N. , Patel T. , Contreras L.M. Discovery and characterization of native deinococcus radiodurans promoters for tunable gene expression . Appl. Environ. Microbiol. 2019 ; 85 : e01356-19 .
Meima R. , Lidstrom M.E. Characterization of the minimal replicon of a cryptic deinococcus radiodurans SARK plasmid and development of versatile Escherichia coli - D. radiodurans shuttle vectors . Appl. Environ. Microbiol. 2000 ; 66 : 3856 – 3867 .
Devigne A. , Mersaoui S. , Bouthier-de-la-Tour C. , Sommer S. , Servant P. The PprA protein is required for accurate cell division of gamma-irradiated Deinococcus radiodurans bacteria . DNA Repair (Amst.) . 2013 ; 12 : 265 – 272 .
Allerston C.K. , Lee S.Y. , Newman J.A. , Schofield C.J. , McHugh P.J. , Gileadi O. The structures of the SNM1A and SNM1B/apollo nuclease domains reveal a potential basis for their distinct DNA processing activities . Nucleic Acids Res. 2015 ; 43 : 11047 – 11060 .
Adachi M. , Hirayama H. , Shimizu R. , Satoh K. , Narumi I. , Kuroki R. Interaction of double-stranded DNA with polymerized PprA protein from Deinococcus radiodurans . Protein Sci. 2014 ; 23 : 1349 – 1358 .
Adachi M. , Shimizu R. , Shibazaki C. , Satoh K. , Fujiwara S. , Arai S. , Narumi I. , Kuroki R. Extended structure of pleiotropic DNA repair-promoting protein PprA from Deinococcus radiodurans . FASEB J. 2019 ; 33 : 3647 – 3658 .
Hirsch A.K.H. , Fischer F.R. , Diederich F. Phosphate recognition in structural biology . Angew. Chem. Int. Ed. Engl. 2007 ; 46 : 338 – 352 .
Yang W. Structure and mechanism for DNA lesion recognition . Cell Res. 2008 ; 18 : 184 – 197 .
Depew D.E. , Wang J.C. Conformational fluctuations of DNA helix . Proc. Natl. Acad. Sci. U.S.A. 1975 ; 72 : 4275 – 4279 .
Pulleyblank D.E. , Shure M. , Tang D. , Vinograd J. , Vosberg H.P. Action of nicking-closing enzyme on supercoiled and nonsupercoiled closed circular DNA: formation of a Boltzmann distribution of topological isomers . Proc. Natl. Acad. Sci. U.S.A. 1975 ; 72 : 4280 – 4284 .
Irobalieva R.N. , Fogg J.M. , Catanese D.J. , Sutthibutpong T. , Chen M. , Barker A.K. , Ludtke S.J. , Harris S.A. , Schmid M.F. , Chiu W. et al. . Structural diversity of supercoiled DNA . Nat. Commun. 2015 ; 6 : 8440 .
McCauley M.J. , Williams M.C. Mechanisms of DNA binding determined in optical tweezers experiments . Biopolymers . 2007 ; 85 : 154 – 168 .
Pyne A.L.B. , Noy A. , Main K.H.S. , Velasco-Berrelleza V. , Piperakis M.M. , Mitchenall L.A. , Cugliandolo F.M. , Beton J.G. , Stevenson C.E.M. , Hoogenboom B.W. et al. . Base-pair resolution analysis of the effect of supercoiling on DNA flexibility and major groove recognition by triplex-forming oligonucleotides . Nat. Commun. 2021 ; 12 : 1053 .
Sutthibutpong T. , Matek C. , Benham C. , Slade G.G. , Noy A. , Laughton C.K. , Doye J.P. , Louis A.A. , Harris S.A Long-range correlations in the mechanics of small DNA circles under topological stress revealed by multi-scale simulation . Nucleic Acids Res. 2016 ; 44 : 9121 – 9130 .
Fogg J.M. , Judge A.K. , Stricker E. , Chan H.L. , Zechiedrich L. Supercoiling and looping promote DNA base accessibility and coordination among distant sites . Nat. Commun. 2021 ; 12 : 5683 .
Matek C. , Ouldridge T.E. , Doyle J.P.K. , Louis A.A. Plectoneme tip bubbles: Coupled denaturation and writhing in supercoiled DNA . Sci. Rep. 2015 ; 5 : 7655 .
Pandey N. , Black B.E. Rapid detection and signaling of DNA damage by PARP-1 . Trends Biochem. Sci. 2021 ; 46 : 744 – 757 .
Eustermann S. , Wu W.-F. , Langelier M.-F. , Yang J.-C. , Easton L.E. , Riccio A.A. , Pascal J.M. , Neuhaus D. Structural basis of detection and signaling of DNA single-strand breaks by Human PARP-1 . Mol. Cell . 2015 ; 60 : 742 – 754 .
Chen X. , Velmurugu Y. , Zheng G. , Park B. , Shim Y. , Kim Y. , Liu L. , Van Houten B. , He C. , Ansari A. et al. . Kinetic gating mechanism of DNA damage recognition by Rad4/XPC . Nat. Commun. 2015 ; 6 : 5849 .
Velmurugu Y. , Chen X. , Slogoff Sevilla P. , Min J.-H. , Ansari A. Twist-open mechanism of DNA damage recognition by the Rad4/XPC nucleotide excision repair complex . Proc. Natl. Acad. Sci. U.S.A. 2016 ; 113 : E2296 – E305 .
Month: | Total Views: |
---|---|
July 2024 | 507 |
August 2024 | 3,243 |
Citing articles via.
Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide
Sign In or Create an Account
This PDF is available to Subscribers Only
For full access to this pdf, sign in to an existing account, or purchase an annual subscription.
Heavier group 14-transition metal π-complex congeners †.
a Fakultät für Chemie, Technische Universität München, Lichtenbergstraße 4, 85748 Garching bei München, Germany E-mail: [email protected]
Since the dawn of organometallic chemistry, transition metal complexes of unsaturated organic molecules, namely π-complexes, have remained a central focus: our thorough understanding of the electronic nature of such species, and their importance in countless reactive processes continues to drive research in their synthesis and utilisation. Since the late 1900s, research regarding the related chemistry for the heavier group 14 elements has become increasingly more fervent. Today, heavier congeners of a vast array of classical π-complexes have been realised, from alkene to arene systems, involving Si, Ge, Sn, and Pb. This has given deeper insights into the bonding observed for these heavier elements, which typically involves a lessened degree of π-bonding and an increased polarisation. This review aims to summarise this field, identifying these disparities, and highlighting areas which we believe may be exciting for future exploration.
Permissions.
T. J. Hadlington, Chem. Soc. Rev. , 2024, Advance Article , DOI: 10.1039/D4CS00497C
This article is licensed under a Creative Commons Attribution 3.0 Unported Licence . You can use material from this article in other publications without requesting further permissions from the RSC, provided that the correct acknowledgement is given.
Read more about how to correctly acknowledge RSC content .
Search articles by author.
This article has not yet been cited.
You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.
All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .
Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.
Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.
Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.
Original Submission Date Received: .
Find support for a specific problem in the support section of our website.
Please let us know what you think of our products and services.
Visit our dedicated information section to learn more about MDPI.
Anomaly detection for charging voltage profiles in battery cells in an energy storage station based on robust principal component analysis, 1. introduction, 2. source and preprocessing of data, 3. anomaly detection process for battery cells, 3.1. the principle of rpca, 3.2. consistency assessment for battery cells, 3.3. sceening and identification, 4. experimental analysis and verification, 4.1. experimental analysis, 4.2. comparison and verification, 4.3. anomaly reasons and analysis, 5. conclusions, author contributions, institutional review board statement, informed consent statement, data availability statement, conflicts of interest.
Click here to enlarge figure
Specification | Value |
---|---|
Battery type | LFP |
Total voltage (V) | 761.6 |
Battery charging termination voltage (V) | 3.65 |
Battery discharge termination voltage (V) | 2.7 |
Nominal voltage (V) | 3.2 |
Nominal capacity (mAh) | 3000 |
Method | The Results on 30 June | The Results on 1 July |
---|---|---|
Average Deviation-3σ | 3, 21, 33, 53, 58, 108 | 3, 21, 33, 53, 58, 108 |
Variance-3σ | 3, 21, 33, 53, 58, 108 | 3, 33, 53, 58, 108 |
Range-3σ | 3, 21, 33, 53, 58, 108 | 3, 21, 33, 53, 58, 108 |
Euclidean Distance-3σ | 3, 21, 33, 53, 58 | 3, 21, 33, 53, 58, 108 |
Signal Energy-3σ | 3, 21, 33, 53, 58 | 3, 21, 33, 53, 58, 108 |
The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
Yu, J.; Guo, Y.; Zhang, W. Anomaly Detection for Charging Voltage Profiles in Battery Cells in an Energy Storage Station Based on Robust Principal Component Analysis. Appl. Sci. 2024 , 14 , 7552. https://doi.org/10.3390/app14177552
Yu J, Guo Y, Zhang W. Anomaly Detection for Charging Voltage Profiles in Battery Cells in an Energy Storage Station Based on Robust Principal Component Analysis. Applied Sciences . 2024; 14(17):7552. https://doi.org/10.3390/app14177552
Yu, Jiaqi, Yanjie Guo, and Wenjie Zhang. 2024. "Anomaly Detection for Charging Voltage Profiles in Battery Cells in an Energy Storage Station Based on Robust Principal Component Analysis" Applied Sciences 14, no. 17: 7552. https://doi.org/10.3390/app14177552
Article access statistics, further information, mdpi initiatives, follow mdpi.
Subscribe to receive issue release notifications and newsletters from MDPI journals
COMMENTS
The introduction looks slightly different depending on whether your paper presents the results of original empirical research or constructs an argument by engaging with a variety of sources. The five steps in this article will help you put together an effective introduction for either type of research paper.
Define your specific research problem and problem statement. Highlight the novelty and contributions of the study. Give an overview of the paper's structure. The research paper introduction can vary in size and structure depending on whether your paper presents the results of original empirical research or is a review paper.
Download Article. 1. Announce your research topic. You can start your introduction with a few sentences which announce the topic of your paper and give an indication of the kind of research questions you will be asking. This is a good way to introduce your readers to your topic and pique their interest.
Research paper introduction is the first section of a research paper that provides an overview of the study, its purpose, and the research question (s) or hypothesis (es) being investigated. It typically includes background information about the topic, a review of previous research in the field, and a statement of the research objectives.
Abstract. An article primarily includes the following sections: introduction, materials and methods, results, discussion, and conclusion. Before writing the introduction, the main steps, the heading and the familiarity level of the readers should be considered. Writing should begin when the experimental system and the equipment are available.
2. Lay a foundation of information already known by presenting findings of other researchers on aspects of the problem you addressed. 3. Indicate the need for more investigation by highlighting a gap in the existing work, showing a need for extension of the work, or creating a research 'niche' that your study fills. 4.
The introduction supplies sufficient background information for the reader to understand and evaluate the experiment you did. It also supplies a rationale for the study. Goals: Present the problem and the proposed solution. Presents nature and scope of the problem investigated. Reviews the pertinent literature to orient the reader.
2. Lay a foundation of information already known by presenting findings of other researchers on aspects of the problem you addressed. 3. Indicate the need for more investigation by highlighting a gap in the existing work, showing a need for extension of the work, or creating a research 'niche' that your study fills. 4.
Step 2: Building a solid foundation with background information. Including background information in your introduction serves two major purposes: It helps to clarify the topic for the reader. It establishes the depth of your research. The approach you take when conveying this information depends on the type of paper.
The introduction of your research paper sets the tone for your research and provides the context for your study. In this article, we will guide you through the process of writing an effective introduction that grabs the reader's attention and captures the essence of your research paper. Understanding the Purpose of a Research Paper Introduction
Author content. Content may be subject to copyright. How to Write a Resear ch Paper Introduction. Step 1: Introduce your topic. The first job of the introdu ction is to tell the reader what your ...
This paragraph should both attract the reader's attention and give them the necessary information about the paper. In any academic paper, the introduction paragraph constitutes 10% of the paper's total word count. For example, if you are preparing a 3,000-word paper, your introduction paragraph should consist of approximately 300 words.
Provide your readers with a road map to help them understand what you will address throughout the research. Be succinct - it is advised that your opening introduction consists of around 8-9 percent of the overall amount of words in your article (for example, 160 words for a 2000 words essay). Make a strong and unambiguous thesis statement.
1- In this paper, I will discuss climate change. Problem: This statement is too broad and vague. It does not provide a clear direction or specific argument. 2- This paper argues that climate change, measured by global average temperature change, is primarily driven by human activities, such as.
Generally speaking, a good research paper introduction includes these parts: 1 Thesis statement. 2 Background context. 3 Niche (research gap) 4 Relevance (how the paper fills that gap) 5 Rationale and motivation. First, a thesis statement is a single sentence that summarizes the main topic of your paper.
Tools for writing a research paper introduction. Now that we've introduced you to the basics of writing a research paper introduction, we'd like to introduce you to QuillBot. At every step of writing your intro, it can help you upgrade your writing skills: Cite sources using the Citation Generator. Avoid plagiarism using the Plagiarism Checker.
Abstract. Ideally, the Introduction is an essential attention grabbing section of a research paper. If written correctly, the Introduction peaks the reader's interest as well as serves as a roadmap for the rest of the paper. An effective Introduction builds off related empirical research and demonstrates a gap in which the current study fills.
The introduction and background sections to a research article are often overlooked and fitted in around the study design. Everyone is understandably keen to write up their method and publish their results. But not only do these sections set the tone and structure for both the article and the study to be described, they also have the potential ...
The introduction serves the purpose of leading the reader from a general subject area to a particular field of research. It establishes the context of the research being conducted by summarizing current understanding and background information about the topic, stating the purpose of the work in the form of the hypothesis, question, or research problem, briefly explaining your rationale ...
Social Science (and Science) original research articles generally follow IMRD: Introduction- Methods-Results-Discussion. Introduction. Introduces topic of article; Presents the "Research Gap"/Statement of Problem article will address; How research presented in the article will solve the problem presented in research gap.
Introduction. Most articles will start with an introductory section, which may be labeled introduction. This section introduces the research study, the thesis statement and why the research being conducted is important. Questions to ask while you read:
Reading the conclusion will help you understand the main points of the article and what the authors are attempting to prove. 3. Read the Introduction Section Now that you have an overview of the article from the abstract and understand the main points the authors are trying to prove from the conclusion, you will want to read the introduction. 4.
The introduction of a research paper includes several key elements: A hook to catch the reader's interest. Relevant background on the topic. Details of your research problem. and your problem statement. A thesis statement or research question. Sometimes an overview of the paper. Frequently asked questions: Writing a research paper.
Accepted February 25, 2023. This paper provides an in-depth introduction to r esearch methods. and discusses numerous aspects r elated to the r esearch process. It. begins with an overview of ...
Editor's note: This is the 23rd article in a series on clinical research by nurses. The series is designed to be used as a resource for nurses to understand the concepts and principles essential to research. Each column will present the concepts that underpin evidence-based practice—from research design to data interpretation.
Results will include books and ebooks, articles, and media like DVDs; A great option when you are just beginning your research process. Use Everything when you want to survey the research landscape for your topic; Library Catalog: Focuses on VU owned items, physical and digital; Results will include books and ebooks, and media like DVDs
Research Article. ANTHROPOLOGY. Share on. Early science and colossal stone engineering in Menga, a Neolithic dolmen (Antequera, Spain) ... INTRODUCTION. Megaliths are structures made of large stones and are found in variety of regions throughout the world. In Late Prehistoric Europe, megalithic monumentality was a widespread phenomenon ...
Introduction. The bacterium Deinococcus radiodurans, along with other species of the Deinococcus genus, are distinguished for their ability to survive high doses of DNA-damaging agents, such as UV-C radiation, ionizing radiation, and desiccation (1, 2).Several factors have been proposed to contribute to the DNA-damage resistance phenotype. Most notably is the atypically high intracellular ...
Since the late 1900s, research regarding the related chemistry for the heavier group 14 elements has become increasingly more fervent. Today, heavier congeners of a vast array of classical π-complexes have been realised, from alkene to arene systems, involving Si, Ge, Sn, and Pb. ... Advance Article , DOI: 10.1039/D4CS00497C This article is ...
In order to solve this problem, this article proposes an anomaly detection method for battery cells based on Robust Principal Component Analysis (RPCA), taking the historical operation and maintenance data of a large-scale battery pack from an energy storage station as the research subject.