Nuclear Magnetic Resonance Solution Structure of DNA Featuring

Article pubs.acs.org/biochemistry

Nuclear Magnetic Resonance Solution Structure of DNA Featuring Clustered 2′-Deoxyribonolactone and 8‑Oxoguanine Lesions Jan Zálešaḱ ,† Jean-François Constant, and Muriel Jourdan* Universite Grenoble Alpes, DCM UMR 5250, F-38000 Grenoble, France CNRS, DCM UMR 5250, F-38000 Grenoble, France S Supporting Information *

ABSTRACT: Ionizing radiation, free radicals, and reactive oxygen species produce hundreds of different DNA lesions. Clustered lesions are typical for ionizing radiation. They compromise the efficiency of the base excision repair (BER) pathway, and as a consequence, they are much more toxic and mutagenic than isolated lesions. Despite their biological relevance, e.g., in cancer radiotherapy and accidental exposure, they are not very well studied from a structural point of view, and while insights provided by structural studies contribute to the understanding of the repair process, only three nuclear magnetic resonance (NMR) studies of DNA containing clusters of lesions were reported. Herein, we report the first NMR solution structure of two DNAs containing a bistranded cluster with the 2′-deoxyribonolactone and 8-oxoguanine lesions. Both DNA duplexes feature a 2′-deoxyribonolactone site in the middle of the sequence of one strand and differ by the relative position of the 8-oxoguanine, staggered 3′ or 5′ side on the complementary strand at a three-nucleotide distance. Depending on its relative position, the repair of the 8-oxoguanine lesion by the base excision repair protein Fpg is either almost complete or inhibited. We found that the structures of the two DNAs containing a bistranded cluster of two lesions are similar and do not deviate very much from the standard B-form. As no obvious structural deformations were observed between the two duplexes, we concluded that the differences in Fpg activity are not due to differences in their global conformation.

E

DNA clustered damage sites frequently contain oxidized nucleobases such as 8-oxoguanine (8-oxo-7,8-dihydro-2′deoxyguanosine), DHT (5,6-dihydroxy-5,6-dihydrothymine), and TG (thymine glycol), nucleobase degradation products, strand breaks, and abasic sites.17−19 Many studies have addressed, using synthetic models, how efficiently these lesions, when organized in cluster, are repaired in vitro,1,2,5,10,20−24 and some have focused on their structural features.25−27 So far, the majority of the reported studies on AP site-containing clusters concern the 2′-deoxyribose abasic site (dR) or its stable tetrahydrofuran analogue. We recently reported the first nuclear magnetic resonance (NMR) structural study of bistranded clusters containing two different combinations of 8-oxoguanine and 2′-deoxyribose lesions for which the cleavage of the 8oxoguanine by the Fpg (formamido-pyrimidine-DNA glycosylase) protein was either quantitative or totally inhibited. The Fpg protein is a bifunctionnal glycosylase. It cleaves the 8oxoguanine N-glycosidic bond to generate an abasic site intermediate, which in turn can be cleaved by the AP-lyase activity. Both bistranded clusters were fairly similar, and we concluded that the dramatic effect on Fpg activity is not due to structural features. Interestingly, our study revealed higher

xposure of DNA to ionizing radiation as well as free radicals and reactive oxygen species generates about one hundred instances of damage, including nucleobases and sugar oxidation and/or fragmentation. DNA lesions induced by ionizing radiation display a spatial distribution much different from that caused by oxygen metabolism, because different types of ionizing radiation deposit their energy very locally. In this context, the production of large amounts of clustered damage sites and direct double-strand break is considered as a signature of the ionizing radiation effects.1−5 A cluster of lesions is defined by the presence of two or more lesions within one or two helical turns of the DNA either on the same strand (tandem lesion) or on opposite strands (bistranded lesions).6,7 The clustered forms of damage are mainly repaired by the base excision repair (BER) pathway,2,5,8−11 and the involvement of either the short or the long-patch subpathway is still not known and depends on the type of lesion and many other factors.12 Additionally, numerous reports have shown that the type of lesions, their relative orientation, and the interlesion distances affect their repair and might lead to very different biological outcomes (reviewed in refs 1−3 and references herein), such as mutation or double-strand breaks. Generally, the clustered DNA lesions compromise the BER efficiency, thus leading to mutation frequencies higher than those of isolated lesions.3,13−16 © 2016 American Chemical Society

Received: April 26, 2016 Revised: June 20, 2016 Published: June 20, 2016 3899

DOI: 10.1021/acs.biochem.6b00396 Biochemistry 2016, 55, 3899−3906

Article

Biochemistry flexibility in the vicinity of the abasic lesion.28 In this paper, we report on the structural features of different clusters containing a 2′-deoxyribonolactone abasic lesion (L) and an 8-oxoguanine. The 2′-deoxyribonolactone (Figure 1a) is an oxidized abasic

cleavage by Fpg of 8-oxoguanine located in symmetrical positions 3 bp from the 2′-deoxyribonolactone position is either quantitative [position (dL)+3] or inhibited [position (dL)−3] (Figure 1c,d). We report the NMR solution structures of these two clusters (Figure 1b) and analyze the perturbations that the cluster of lesions causes on structural and dynamic properties of the DNA. The other symmetrical positions of the 8-oxoguanine (positions +5 and −5 and positions +1 and −1) were not studied from a structural point of view as the activity of the Fpg is efficient and similar for positions +5 and −5 and very weak for positions +1 and −1.

■

MATERIALS AND METHODS Enzyme Cleavage Assay. The cleavage activity of the E. coli Fpg was measured by monitoring the conversion of 32Plabeled oligonucleotide duplexes containing the 2′-deoxyribonolactone and 8-oxoguanine lesions at defined positions (see panels c and d of Figure 1) into cleavage products. The 31-mer duplexes were prepared by hybridizing the 32P-labeled oligonucleotides containing the 8-oxoguanine with a 10% molar excess of the nonlabeled complementary strand containing the 2′-deoxyribonolactone lesion at room temperature. The optimal Fpg protein concentration was determined in preliminary experiments with DNA duplexes containing a single 8-oxoguanine lesion at position 0 (Figure 1d) to obtain 25−30% of cleavage products in 20 min at 37 °C. The labeled oligonucleotide duplex containing the bistranded cluster (25 nM) was incubated with the Fpg protein (16.5 nM) for 20 min at 37 °C (final reaction volume of 10 μL). The enzymatic reaction was quenched by adding 10 μL of stop solution [95% formamide, 20 mM EDTA (pH 7), 0.025% xylene cyanol, and 0.025% bromophenol blue]. The reaction products were then separated by polyacrylamide gel electrophoresis under denaturing conditions [7 M urea, (19:1) acrylamide:bis(acrylamide), and 100 mM HEPES (pH 7)] at 50 W for 1.5 h. The gels (Figure 1c) were analyzed using a Typhoon 9410 instrument (GE Healthcare), and the incision products were quantified with Image Quant TL. The percentage of enzymatic cleavage corresponds to the ratio of the cleaved DNA strand over the sum of the intact DNA strand and cleaved strand. All experiments were repeated thrice. Melting Temperature Measured by Circular Dichroism. Circular dichroism melting temperature experiments were conducted using a JASCO-810 CD spectrophotometer equipped with a Peltier temperature controller. The temperature was increased from 5 to 80 °C at a rate of 0.5 °C/min, and the data were recorded at 252 nm. The DNA sample concentration was 55 μM in 20 mM phosphate buffer (pH 6) and 100 mM NaCl. The melting temperature was determined using the first derivative of the experimental melting curve. NMR Studies. Sample Preparation. The oligonucleotides containing the 8-oxoguanine were purchased from Eurogentec (Liège, Belgium). The 2′-deoxyribonolactone-containing oligonucleotides were synthesized according to the method described previously.35 Briefly, the phosphoramidite synthon of the photoactivatable precursor 7-nitroindole (dNi) was incorporated by standard solid-phase automated synthesis in the sequence d(CGCTCdNiCACGC). The oligonucleotide was purified by high-performance liquid chromatography and ion exchange chromatography. The duplexes (dNi)+3 and (dNi)−3 were formed by mixing at room temperature equimolar amounts of d(CGCTCdNiCACGC) with d(GCG-

Figure 1. (a) Chemical structures of the 2′-deoxyribonolactone (L) and 8-oxoguanine (8G) lesions. (b) Sequence and numbering of the 11-mer duplexes (dL)+3 and (dL)−3 used in the NMR study. (c) Qualitative cleavage assay (analyzed by denaturing polyacrylamide gel electrophoresis) showing the cleavage efficiency of the E. coli Fpg protein at the 8-oxoguanine position incorporated in 5′-32P-labeled DNA sequences shown in panel d (31-mer). Almost no cleavage (70%) is observed for position +3 [in terms of experiment details, the 32Plabeled (0.05 pmol) DNA was incubated in the presence of Fpg protein (35 nM) for 20 min at 37 °C (final reaction volume of 10 μL)]. (d) Sequence of the seven different 31-mers used for the enzymatic cleavage assay. The L lesion is at a fixed position, and 8oxoguanine is at various positions (−5 to +5) on the opposite strand, marked by an asterisk. The 11-mer sequence used for NMR studies is shown in bold.

site resulting from hydrogen abstraction at the anomeric position of nucleotide by reactive radical species generated by drugs (neocarzinostatin or TMPyP)29−31 or γ and UV irradiation.32−34 To date, there has been a serious lack of information about clusters containing this abasic lesion. A recent study of a bistranded cluster comprising combinations of 8-oxoguanine and 2′-deoxyribonolactone revealed an increased mutagenic potential compared to that of isolated lesions16 (more mutations occur than for isolated lesions in wild-type Escherichia coli or fpg, mutY, and f pgmutY deficient E. coli strains). It has also been shown that the presence of the 8oxoguanine does not affect the repair of the 2′-deoxyribonolactone lesion by BER enzymes. Herein, we show that the 3900

DOI: 10.1021/acs.biochem.6b00396 Biochemistry 2016, 55, 3899−3906

Article

Biochemistry TGGGA8GCG) and d(GC8GTGGGAGCG), respectively. The dNi precursor was quantitatively converted into the 2′deoxyribonolactone by illuminating the duplexes at 4 °C (to prevent strand cleavage by β-elimination) using a 200 W Hg/ Xe lamp (Oriel Instrument, Stratford, CT). Radiation wavelengths below 300 nm were filtered by using a 2 M KNO3 filter.35 The conversion of dNi into 2′-deoxyribonolactone and the degradation of 2′-deoxyribonolactone were controlled by alkali treatment and gel electrophoresis.36 The dNi conversion was nearly complete after illumination for 90 min, and no degradation products were detected. The samples were then lyophilized and resuspended in 99.99% D2O. For NMR investigation of exchangeable protons, the oligomers were dissolved in a 90% H2O/10% D2O solvent (v/v). The final duplex concentration was 0.85 mM in 20 mM sodium phosphate buffer (pH 6) and 100 mM NaCl for both (dL)+3 and (dL)−3. NMR Experiments. All 1H NMR experiments were conducted with a Varian Unity Plus 500 MHz spectrometer (NMR-ICMG Platform, Grenoble, France). Chemical shifts were calibrated relative to sodium 3-(trimethylsilyl)propionate2,2,3,3-d4 (TSP-d4). Data were processed with VNMR software (Varian) and analyzed with CCPNMR.37 NOESY spectra with mixing times of 75, 100, 150, and 250 ms and TOCSY spectra with mixing times of 40 and 60 ms were recorded in D2O buffer at 10 °C as previously described.28 NOESY spectra in 90% H2O (mixing time of 350 ms) were recorded at 5 °C, and water was suppressed using sculpting gradients. Distance and Dihedral Angle Restraints. The NOE intensities measured from the NOESY spectra in D2O at three different mixing times (75, 100, and 150 ms) were converted into distance restraints using the CCPNMR tool and calibrated to an internal standard (cytosine H5−H6 = 2.45 Å). The lower and upper boundaries were set to 0.8 and 1.2 multiples of calculated distances, respectively, thus allowing the calculations to handle the experimental error. A qualitative determination of the JH1′−H2′, JH2″−H3′, and JH3′−H4′ coupling constants was gained from analysis of the DQFCOSY crosspeak intensities and used to derive pseudorotation angles P deduced from these values (except for 2′-deoxyribonolactone site residue L6). A P value in the range 165−180° was used for the C2′-endo conformation, and the range of 80−115° was used for the O4′-endo conformation. The interproton distances and P values were then processed using the Antechamber module (AMBER version 10.1, University of California, San Francisco, CA) to generate the distances and sugar pucker restraints. Molecular Modeling. The B-DNA duplexes were built by using InsightII software, and calculations were performed as previously described28 with AMBER version 10.2.38 The AMBER parameters used to define the 8-oxoguanine library were taken from the Bryce group web page [Manchester University, North Manchester, IN (http://www.pharmacy. manchester.ac.uk/bryce/amber#nuc)]. A library for the 2′deoxyribonolactone was created from the RESP ESP chargederived server developed by Dupradeau and collaborators (University of Picardie Jules Verne, Amiens, France).39 Four random initial structures were prepared using 3−5 ps unrestrained molecular dynamics at elevated temperatures ranging from 600 to 1000 K. Each of these structures and one B-DNA-shaped structure were then subjected to a 200 ps restrained simulated annealing protocol with implicit solvent (Born solvation model). The restraints used for (dL)+3 and

(dL)−3 corresponded to 356 and 295 experimental interproton distances and 105 and 110 sugar dihedral angles, respectively. On the basis of the experimental data, we imposed 87 backbone torsion angles (maintaining the right-handed character of the helix) and Watson−Crick base pairing restraints (31 distances and 19 angles). No Watson−Crick restraints were imposed for the 8-oxoguanine-containing base pairs (8G14·C9 or 8G20·C3) or the 2′-deoxyribonolactone region (C5·G18, L6·G17, and C7· G16). The five final structures were averaged and solvated in a periodic TIP3P truncated octahedron water box that extended 4.5 Å from any solute atom. Once the system was fully equilibrated, a dynamics of 1 ns was performed. The 10 structures with the lowest energy over the last 40 ps were analyzed, averaged, and minimized. Properties of the system during dynamics (potential, kinetic, and total energies, volume, pressure, and others) were verified as reasonable using the ptraj module of AMBER. The helical parameters were analyzed using Curves+.40 Data Deposition. The coordinates of the structures have been deposited in the Protein Data Bank as entries 5HQF for (dL)+3 and 5HQQ for (dL)−3. NMR assignments have been deposited in the BioMagResBank as entries 30015 for (dL)+3 and 30016 for (dL)−3.

■

RESULTS

In Vitro Activity of Fpg on Bistranded Cluster Lesions. In vitro enzymatic cleavage assays of the bistranded cluster lesions 31-mer DNA (Figure 1c,d) were performed with the E. coli Fpg protein. As shown in Figure 1c, the cleavage of the 8oxoguanine at positions +5 and −5 is quite efficient and similar. For the symmetrical positions +3 and −3, a strikingly different behavior is observed, with no cleavage at position −3 (70%). For positions +1, 0, and −1, the cleavage is strongly inhibited. The NMR study was performed on shorter oligonucleotides (dL)+3 and (dL)−3 (11-mer) (Figure 1b) consisting of the central 11 bp of the 31mer and comprising the clusters with 8-oxoguanine at positions +3 and −3, respectively. Melting Experiments. The melting temperatures of the DNA containing clusters (dL)+3 and (dL)−3 were measured by circular dichroism (Table 1). The presence of the two lesions produced strong destabilizing effects of 26 and 29 °C for (dL)+3 and (dL)−3, respectively, compared to the undamaged duplex. The melting temperature of a duplex containing a single 2′-deoxyribonolactone is ∼23 °C lower than that of the undamaged duplex. Table 1. Melting Temperatures for DNA Duplexes Measured by CD DNA duplexa CGCTCCC ACGC GCGAGGGTGCG CG CT CCCACGC GC8GAGGGTGCG CGCTCCCA CGC GCGAGGGT8GCG

Tm (°C) b

70

64b 61b

DNA duplexa

Tm (°C)

CGCTCLCACGC GCGAGGGTGCG CG C TCLCACGC GC8GAGG GTGCG CGCTCLCA CGC GCGAGGGT 8GCG

47 44 41

a The DNA concentration was 55 μM in 20 mM phosphate buffer (pH 6) and 100 mM NaCl. The observed standard deviation is ±1 °C. b Value taken from ref 28.

3901

DOI: 10.1021/acs.biochem.6b00396 Biochemistry 2016, 55, 3899−3906

Article

Biochemistry

in (dL)−3, the intrahelical position of 8G14 is demonstrated by NOE signals between T15 H6 and 8G14 deoxyribose protons (Figure 2). The unpaired base G17 was also found to be intrahelical as NOE cross-peaks between G18 H8 and G17 sugar protons and between G17 H8 and G16 sugar protons were found in both (dL)+3 and (dL)−3 duplexes. The cross-peak intensities observed for all of the nucleobases, including 8G, are consistent with an anti conformation of the nucleotides and with a global B-like DNA conformation of the duplexes. The N-glycosidic bond conformation for 8oxoguanine in (dL)+3 was evidenced by NOEs between base proton C21 H6 and deoxyribose protons 8G20 H1′, H2′, and H2″. In the case of (dL)−3, NOEs were measured between T15 H6 and 8G14 H1′, H2′, and H2″. Moreover, the anti conformation of the 8G nucleotide brings the H1′ and H2″ protons of the adjacent 5′-nucleotide into the proximity of the 8 G carbonyl, causing a significant downfield shift of their signals [A19 H1′, 6.35 ppm; A19 H2″, 3.20 ppm for (dL)+3; C13 H1′, 6.19 ppm; C13 H2″, 2.73 ppm for (dL)−3]. As seen previously,42 the H2′ protons of 8G20 and 8G14 were found downfield from the H2″ protons and were unambiguously assigned thanks to strong NOE and COSY cross-peaks between H1′ and H2′. Resonance Assignments of Exchangeable Protons. The assignment of the exchangeable imino and amino proton for both duplexes was achieved by analysis of the one- and twodimensional NOESY spectra recorded in 90% H2O at 5 °C. The non-hydrogen-bound imino protons of G17 facing the abasic site were found upfield at 10.15 and 10.12 ppm for (dL)+3 and (dL)−3, respectively, in accordance with the chemical shift expected for an unpaired base facing an abasic site in a similar sequence context43 and consistent with our previous study.28 The other chemical shifts and observed NOEs were indicative of standard Watson−Crick base pairing: each thymine imino proton displays a strong NOE cross-peak to the H2 proton of its partner adenine, and the guanine imino proton exhibits a cross-peak to the amino proton of its cytosine partner (Figure 3). However, weaker NOE cross-peaks for C7·G16 and C5·G18 base pairs were observed. Together with the upfield shift of the respective G18 N1H and G16 N1H protons (from

Resonance Assignments of Nonexchangeable Protons. Proton resonances for duplexes (dL)+3 and (dL)−3 were assigned from standard two-dimensional experiments using a classical sequential assignment strategy for right-handed DNA duplexes.41 The 1H resonance assignments are available in Table S1 for (dL)+3 and Table S2 for (dL)−3. Some H5′/ H5″ protons could not be assigned because of severe overlap. We used 150 ms NOESY spectra to establish sequential connectivities between base protons (H6 and H8) and deoxyribose protons (H1′, H2′/H2″, H3′, H4′, and most of H5′/H5″). These base-to-sugar NOE connectivities were disrupted between C5 and C7 because of the absence of a base moiety in the L6 residue. However, the signals observed between C7 H6 and L6 deoxyribonolactone protons H2′ and H2″ were weaker than those observed in standard B-DNA, while medium NOE signals were detected between C7 H6 and L6 H3′ (Table 2 and Figure 2). These NOEs are particularly Table 2. List of NOE Interactions Observed between the Protons of 2′-Deoxyribonolactone L6 and the Flanking Bases C5 and C7 in (dL)+3 and (dL)−3 DNA Duplexesa (dL)+3 C7 C7 C7 C7

H5−L6 H6−L6 H6−L6 H6−L6

H3′ (m) H2′ (w) H2″ (w) H3′ (m)

(dL)−3 C7 H6−L6 H2′ (w) C7 H6−L6 H2″ (w) C7 H6−L6 H3′ (m)

a

Medium (m) and weak (m) NOEs correspond to distances of 3.0− 4.0 and 4.0−5.0 Å, respectively.

crucial because they are used as restraints in molecular dynamics to assess the mutual positions of C5, L6, and C7 residues. They notably indicate that residue L6 is only partially located inside the helix and that a structural perturbation occurs around the lesion. Because of the absence of the H8 proton in the 8-oxoguanine base, the base-to-sugar NOE connections were also interrupted between 8G20 and A19 for (dL)+3 and between 8G14 and C13 for (dL)−3. The strong NOE signals between C21 H6 and 8G20 deoxyribose H2″ and the medium signals between C21 H6 and 8G20 H2′ and H1′ reveal that 8oxoguanine 8G20 is intrahelical in the (dL)+3 duplex. Similarly

Figure 2. Expansion of the 250 ms NOESY spectra at 10 °C showing the correlations between the C5 and C7 bases flanking the deoxyribonolactone site and between the 8-oxoguanine and the base located on its 3′ side [C21 in (dL)+3 and T15 in (dL)−3]. Intermolecular NOEs between C7 and L6, C21 and 8G20, and T15 and 8G14 are denoted by arrows. 3902

DOI: 10.1021/acs.biochem.6b00396 Biochemistry 2016, 55, 3899−3906

Article

Biochemistry

Figure 3. Expansion of the imino and amino regions of NOESY spectra at 5 °C for the (a) (dL)+3 duplex and (b) (dL)−3 duplex. G17 resonating upfield is not shown. The left panel shows the imino proton region, showing NOEs between guanine N1H and thymine N3H. The right panel shows NOEs between guanine N1H and cytosine N4H amino protons and between thymine N3H and adenine H2 protons.

Figure 4. Superimposition of the 10 lowest-energy structures emerging from the restrained molecular dynamics: (a) (dL)+3 duplex and (b) (dL)−3 duplex. The 2′-deoxyribonolactone is colored orange and the 8-oxoguanine green.

for H3′−H4′ and H2″−H3′ (very small JH3′−H4′ and JH2″−H3′ coupling constants), but a characteristic H1′−H2′ pattern along with strong intensities in the NOESY spectra favors a C2′-endo conformation. A large JH3′−H4′ coupling constant and a small but detectable JH2″−H3′ coupling constant are consistent with the O4′-endo conformation. For the L6 2′-deoxyribonolactone sugar, no dihedral angle restraints were used given the low intensity of the cross-peaks.

12.69 to 12.92 ppm), this may reflect a lower stability of these pairs adjacent to the 2′-deoxyribonolactone lesion. The N7H exchangeable proton of 8G20 was found at 9.78 ppm for (dL)+3 displaying a NOE with C21 NH2. For (dL)−3, N7H of 8 G14 displayed a similar chemical shift (10.18 ppm). Sugar Conformations. The sugar conformations were assessed from the DQFCOSY spectra and found to be in the O4′-endo to C2′-endo range, which is consistent with an S-type conformation sugar. No significant cross-peaks were detected 3903

DOI: 10.1021/acs.biochem.6b00396 Biochemistry 2016, 55, 3899−3906

Article

Biochemistry Molecular Dynamics and Structural Features. To assess the structure of (dL)+3 and (dL)−3 DNA duplexes, we performed 1 ns restrained molecular dynamics simulations with experimental restraints in explicit water according to a previously reported protocol.28 It consists of a simulated annealing of the duplexes under restraints applied to five different randomized DNA structures. The distances were then refined to ensure a structural convergence, and the dynamics were performed again with these new distances, until the rootmean-square deviations among the five structures leveled off. The five final structures were averaged and subsequently submitted to a 1 ns dynamics simulation in explicit water. Figure 4 displays an ensemble of the 10 structures with the lowest energy over the last 40 ps of the dynamics for (dL)+3 and (dL)−3. Duplexes (dL)+3 and (dL)−3 adopt a B-DNA conformation and display regular Watson−Crick base pairing. The root-mean-square deviations calculated for all atoms range from 0.32 to 0.70 Å for (dL)+3 and from 0.42 to 0.65 Å for (dL)−3. This indicates that the restraint molecular dynamics converged to well-defined structures. These structures were then averaged for further detailed analysis. On the whole, the averaged structures are in good agreement with the NMR experimental data as illustrated by several quality indexes reported in Table 3.

Figure 5. Detailed view of Watson−Crick pairing of the (a) 8G20·C3 and (b) 8G14·C9 base pairs for (dL)+3 and (dL)−3, respectively, viewed along the helical axis (left) and from the minor groove (right). Detailed structures of the abasic site regions of (c) (dL)+3 and (d) (dL)−3. 2′-Deoxyribonolactone L6 and the unpaired G17 nucleotide are colored for the sake of clarity (C, green; O, red; P, orange; N, blue; H, white).

Table 3. Summary of NMR Violations, Deviations from Ideal Covalent Geometry, and van der Waals Energies for the Averaged (dL)+3 and (dL)−3 Structures no. of violations of NMR restraints distances of >0.2 Å angles of >10° root-mean-square deviation from ideal covalent geometry bond lengths (Å) bond angles (deg) energyvan der Waals (kcal mol−1)

(dL)+3

(dL)−3

3 14

4 14

0.012 2.6 −330.4

0.011 2.6 −334.1

thymine glycol (reviewed in refs 2, 3, and 5). It has been established that these clusters of lesions are more difficult to repair than isolated ones. They compromise the base excision repair pathway, and as a consequence, their lifetime is increased compared to that of isolated lesions.44 Consequently, the lesions are more likely to persist and induce deleterious mutations during replication and transcription steps. It has been shown in E. coli, yeast, and mammalian cells that clusters of lesions can cause very different biological outcomes in terms of mutagenesis and repair inhibition, depending on the type of lesions, the interlesion distance, and their relative orientation. With regard to the clusters containing a 2′-deoxyribonolactone and an 8-oxoguanine, knowledge is very scarce as there has been only one recent report by Lomax and co-workers.16 This cluster in which one of the lesions (the 2′-deoxyribonolactone) is processed by the long-patch BER pathway poses an additionnal challenge to the cell’s repair capacity. Using an E. coli plasmid-based assay, the authors showed that the extent of repair of the 2′-deoxyribonolactone lesion is reduced compared to the extent of repair of a 2′-deoxyribose or a tetrahydrofuran, and that no double-strand break occurs during the repair of the lesion. In addition, their results demonstrated that the proximity of the two lesions (lesion 5 or 1 bp away) results in an increased mutagenic potential of both lesions. Unfortunately, positions +3 and −3 were not investigated, and considering our own results on such clusters, we herein report the first NMR structure of such bistranded clusters. The thermal stability of the clusters was investigated by circular dichroism thermal denaturation experiments. Our results show that a single 2′-deoxyribonolactone affects the stability of the duplex by 23 °C. This is consistent with our previous studies showing that a THF or a 2′-deoxyribose destabilized a duplex by 23 °C in the same sequence context.28 Duplexes (dL)+3 and (dL)−3 are destabilized by 26 and 29 °C, respectively, when the lesions are in a cluster, combining the destabilization of the 8-oxoguanine:C base pair and the 2′deoxyribonolactone baseless site.

Figure 5 displays a close-up view of 8G20·C3 and 8G14·C9 base pairs and of the abasic site region in the averaged structure of (dL)+3 and (dL)−3 duplexes. It shows that the presence of the two lesions does not induce any major distortion. For the (dL)+3 duplex, 2′-deoxyribonolactone residue L6 is not fully intrahelical but is somewhat pointing toward the minor groove. For (dL)−3, residue L6 is almost aligned with the phosphodiester backbone. In both duplexes, the L6 sugar moiety adopts a C3′-exo-like conformation. In addition, the C5· G18 and C7·G16 base pair parameters (bond lengths and angles) are characteristic of a Watson−Crick base pairing in which significant stacking interactions are observed among G16, G17, and G18. However, a slight distortion of the C5·G18 base pair is noticed in (dL)−3 (propeller twist of −20°). The 8oxoguanine nucleotide conformation is anti in both duplexes with χ angles values of −88° for (dL)+3 and −117° for (dL)− 3, and 8-oxoguanine is well paired with its complementary C3 or C9 base (Figure 5).

■

DISCUSSION To date, there have been numerous reports focusing on in vitro and in vivo repair of clusters containing various combinations of forms of damage such as 2′-deoxyribose, tetrahydrofuran, single-strand break, 8-oxoguanine, dihydrothymine, and 3904

DOI: 10.1021/acs.biochem.6b00396 Biochemistry 2016, 55, 3899−3906

Article

Biochemistry

Biology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany.

Prior to the NMR study, an enzymatic cleavage assay was performed to assess the influence of the position of the two lesions on the cleavage efficiency of the 8-oxoguanine by the Fpg. A strong cleavage of the 8-oxoguanine was observed in (dL)+3, whereas almost no cleavage occurred for (dL)−3 (Figure 1c). These results are similar to those we observed for a cluster of 2′-deoxyribose and 8-oxoguanine with the same sequence and for which we determined the NMR structures.28 Previous structural studies of short DNA duplexes containing a single 8-oxoguanine42,45 or a single 2′-deoxyribonolactone lesion43 showed that these lesions do not induce dramatic structural distortions. These findings show that when associated in clusters, these two lesions do not induce severe structural changes either. The (dL)+3 and (dL)−3 duplexes adopt an overall B-form conformation and have similar structural features. No striking deformations of the DNA backbone were noticed as a result of the proximity of the lesions. The damaged residues remain globally intrahelical. Nevertheless, the 8-oxoguanine:C base pair showed an increased flexibility during the dynamics compared to that of standard Watson−Crick base pairs. This is consistent with the known reduced stability of these base pairs.42,46,47 The Watson−Crick pairing of the bases flanking the 2′-deoxyribonolactone is quite stable, and by comparison with our previous report on identical clusters containing a 2′-deoxyribose,28 the structure appears to be more rigid. The transient H-bond detected over the dynamics between G17 and C5 for duplexes containing a cluster of 8oxoguanine and 2′-deoxyribose is not detected in the case of (dL)+3 and (dL)−3. In conclusion, we determined the structures in solution of two duplexes containing different clusters of 8-oxoguanine and 2′-deoxyribonolactone. This provides new structural information about clustered DNA lesions containing 2′-deoxyribonolactone that have been sparsely studied. The fact that (dL)+3 and (dL)−3 are processed differently by Fpg despite their structural similarity is in agreement with our recent studies28 suggesting that structural features might be less important than the sequence specific changes in the kinetics of base pair opening, the thermodynamic profile of flipping the 8oxoguanine out of the helix, and also the desolvation of the protein upon complex formation, which is thought to play an important role in adjusting the complex in a catalytically active state.47

■

Funding

The authors thank the French Agency of National Recherche through the LABEX ARCANE (ANR-11-LABX-0003-01) and the European Community (Lifelong Learning program) for financial support to J.Z. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Jérome Blu for the nitroindole phosphoramidite synthesis and Dr. Rém y Lartia for the synthesis of oligonucleotides. We also thank Sébastien Morin and Pierre Girard for their help with some software setup. The authors acknowledge support from the ICMG FR 2607 Chemistry Nanobio platform. The figures were produced using the UCSF Chimera package freely available from the Resource for Biocomputing, Visualization and Informatics at the University of California, San Francisco.48

■

(1) Magnander, and Elmroth, K. (2012) Biological consequences of formation and repair of complex DNA damage. Cancer Lett. 327, 90− 96. (2) Eccles, L. J., O’Neill, P., and Lomax, M. E. (2011) Delayed repair of radiation induced clustered DNA damage: Friend or foe? Mutat. Res., Fundam. Mol. Mech. Mutagen. 711, 134−141. (3) Sage, E., and Harrison, L. (2011) Clustered DNA lesion repair in eukaryotes: Relevance to mutagenesis and cell survival. Mutat. Res., Fundam. Mol. Mech. Mutagen. 711, 123−133. (4) Okayasu, R. (2012) Repair of DNA damage induced by accelerated heavy ions-a mini review. Int. J. Cancer 130, 991−1000. (5) Georgakilas, A. G., O’Neill, P., and Stewart, R. D. (2013) Induction and repair of clustered DNA lesions: what do we know so far? Radiat. Res. 180, 100−109. (6) Goodhead, D. T. (1994) Initial events in the cellular effects of ionizing radiations: clustered damage in DNA. Int. J. Radiat. Biol. 65, 7−17. (7) Ward, J. F. (1994) The complexity of DNA damage: relevance to biological consequences. Int. J. Radiat. Biol. 66, 427−432. (8) Shikazono, N., Noguchi, M., Fujii, K., Urushibara, A., and Yokoya, A. (2009) The Yield, processing, and biological consequences of custered DNA camage induced by ionizing radiation. J. Radiat. Res. 50, 27−36. (9) Blaisdell, J. O., and Wallace, S. S. (2001) Abortive base-excision repair of radiation-induced clustered DNA lesions in Escherichia coli. Proc. Natl. Acad. Sci. U. S. A. 98, 7426−7430. (10) D’Souza, D. I., and Harrison, L. (2003) Repair of clustered uracil DNA damages in Escherichia coli. Nucleic Acids Res. 31, 4573− 4581. (11) Lomax, M. E., Cunniffe, S., and O’Neill, P. (2004) Efficiency of repair of an abasic site within DNA clustered damage sites by mammalian cell nuclear extracts. Biochemistry 43, 11017−11026. (12) Fortini, P., and Dogliotti, A. (2007) Base damage and singlestrand break repair: mechanisms and functional significance of shortand long-patch repair subpathways. DNA Repair 6, 398−409. (13) Pearson, C. G., Shikazono, N., Tacker, J., and O’Neill, P. (2004) Enhanced mutagenic potential of 8-oxo-7,8-dihydroguanine when present within a clustered DNA damage site. Nucleic Acids Res. 32, 263−270. (14) Budworth, H., Dianova, II, Podust, V. N., and Dianov, G. L. (2002) Repair of clustered DNA lesions. SEQUENCE-SPECIFIC INHIBITION OF LONG-PATCH BASE EXCISION REPAIR BY 8OXOGUANINE. J. Biol. Chem. 277, 21300−21305.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biochem.6b00396. Tables of chemical shifts (Tables 1 and 2), part of the 250 ms NOESY and DQF COSY spectra for (dL)+3 (Figures S1 and S2), and part of the 250 ms NOESY and DQF COSY spectra for (dL)−3 (Figures S3−24) (PDF)

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Telephone: +33 4 56 52 08 39. Fax: +33 4 56 52 08 52. Present Address †

J.Z.: Institute of Molecular Biotechnology, Austrian Academy of Sciences, Vienna, Austria, and Center for Structural Systems 3905

DOI: 10.1021/acs.biochem.6b00396 Biochemistry 2016, 55, 3899−3906

Article

Biochemistry

the Help of Radiation-Chemical Methods. Angew. Chem., Int. Ed. Engl. 30, 1229−1253. (34) Urata, H., and Akagi, M. (1991) Photo-induced formation of the 2-deoxyribonolactone-containing nucleotide for d(ApCpA); effects of neighboring bases and modification of deoxycytidine. Nucleic Acids Res. 19, 1773−1778. (35) Kotera, M., Bourdat, A. G., Defrancq, E., and Lhomme, J. (1998) A highly efficient synthesis of oligodeoxyribonucleotides containing the 2′-deoxyribonolactone lesion. J. Am. Chem. Soc. 120, 11810−11811. (36) Roupioz, Y., Kotera, M., and Lhomme, J. (2002) Chemistry of the 2-deoxyribonolactone lesion in oligonucleotides: Cleavage kinetics and products analysis. J. Am. Chem. Soc. 124, 9129−9135. (37) Vranken, W. F., Boucher, W., Stevens, T. J., Fogh, R. H., Pajon, A., Llinas, M., Ulrich, E. L., Markley, J. L., Ionides, J., and Laue, E. D. (2005) The CCPN data model for NMR spectroscopy: development of a software pipeline. Proteins: Struct., Funct., Genet. 59, 687−696. (38) Case, D. A., Darden, T. A., Cheatham, T. E., Simmerling, C. L., Wang, J., Duke, R. E., et al. (2008) AMBER 10, University of California, San Francisco. (39) Dupradeau, F.-Y., Pigache, A., Zaffran, T., Savineau, C., Lelong, R., Grivel, N., Lelong, D., Rosanski, W., and Cieplak, P. (2010) The R.E.D. tools: Advances in RESP and ESP charge derivation and force field library building. Phys. Chem. Chem. Phys. 12, 7821−7839. (40) Lavery, R., Moakher, M., Maddocks, J. H., Petkeviciute, D., and Zakrzewska, K. (2009) Conformational analysis of nucleic acids revisited: Curves+. Nucleic Acids Res. 37, 5917−5929. (41) Wüthrich, K. (1986) Resonance assignments and structure determination in nucleic acid. In NMR of Proteins and Nucleic Acids, pp 203, Wiley-Interscience, New York. (42) Oda, Y., Uesugi, S., Ikehara, M., Nishimura, S., Kawase, Y., Ishikawa, H., Inoue, H., and Ohtsuka, E. (1991) NMR studies of a DNA containing 8-hydroxydeoxyguanosine. Nucleic Acids Res. 19, 1407−1412. (43) Jourdan, M., Garcia, J., Defrancq, E., Kotera, M., and Lhomme, J. (1999) 2′-Deoxyribonolactone Lesion in DNA: Refined Solution Structure Determined by Nuclear Magnetic Resonance and Molecular Modeling. Biochemistry 38, 3985−3995. (44) Lomax, M. E., Cunniffe, S., and O’Neill, P. (2004) 8-OxoG retards the activity of the ligase III/XRCC1 complex during the repair of a single − strand break, when present within a clustered DNA damage site. DNA Repair 3, 289−299. (45) Lipscomb, L. A., Peek, M. E., Morningstar, M. L., Verghis, S. M., Miller, E. M., Rich, A., Essigmann, J. M., and Williams, L. D. (1995) Xray structure of a DNA 5′-d(CGC-oxoG-AATTCGCG) decamer containing 7,8-dihydro-8-oxoguanine. Proc. Natl. Acad. Sci. U. S. A. 92, 719−723. (46) Singh, S. K., Szulik, M. W., Ganguly, M., Khutsishvili, Stone, M. P., Marky, L. A., and Gold, B. (2011) Characterization of DNA with an 8-oxoguanine modification. Nucleic Acids Res. 39, 6789−6801. (47) Kuznetsov, N. A., Vorobjev, Y. N., Krasnoperov, L. N., and Fedorova, O. S. (2012) Thermodynamics of the multi-stage DNA lesion recognition and repair by formamidopyrimidine-DNA glycosylase using pyrrolocytosine fluorescence-stopped-flow pre-steadystate kinetics. Nucleic Acids Res. 40, 7384−7392. (48) Pettersen, E. F., Goddard, T. D., Huang, C. C., Couch, G. S., Greenblatt, D. M., Meng, E. C., and Ferrin, T. E. (2004) UCSF Chimera–a visualization system for exploratory research and analysis. J. Comput. Chem. 25, 1605−1612.

(15) Eccles, L. J., Lomax, M. E., and O’Neill, P. O. (2010) Hierarchy of lesion processing governs the repair, double-strand break formation and mutability of three-lesion clustered DNA damage. Nucleic Acids Res. 38, 1123−1134. (16) Cunniffe, S., O’Neill, P. O., Greenberg, M. M., and Lomax, M. E. (2014) Reduced repair capacity of a DNA clustered damage site comprised of 8-oxo-7,8-dihydro-2′-deoxyguanosine and 2-deoxyribonolactone results in an increased mutagenic potential of these lesions. Mutat. Res., Fundam. Mol. Mech. Mutagen. 762, 32−39. (17) Pouget, J.-P., Frelon, S., Ravanat, J.-L., Testard, I., Odin, F., and Cadet, J. (2002) Formation of modified DNA bases in cells exposed either to gamma radiation or to high-LET particles. Radiat. Res. 157, 589−595. (18) Cadet, J., Douki, T., and Ravanat, J.-L. (2008) Oxidatively generated damage to the guanine moiety of DNA: mechanic aspects and formation in cells. Acc. Chem. Res. 41, 1075−1083. (19) Frelon, S., Douki, T., Ravanat, J.-L., Pouget, J. P., Tornabene, C., and Cadet, J. (2000) High-performance liquid chromatographytandem mass spectrometry measurement of radiation-induced base damage to isolated and cellular DNA. Chem. Res. Toxicol. 13, 1002− 1010. (20) Blaisdell, J. O., and Wallace, S. S. (2001) Abortive base-excision repair of radiation-induced clustered DNA lesions in Escherichia coli. Proc. Natl. Acad. Sci. U. S. A. 98, 7426−7430. (21) Harrison, L., Brame, K. L., Geltz, L. E., and Landry, A. M. (2006) Closely opposed apurinic/apyrimidinic sites are converted to double strand breaks in Escherichia coli even in the absence of exonuclease III, endonuclease IV, nucleotide excision repair and AP lyase cleavage. DNA Repair 5, 324−335. (22) Gulston, M., de Lara, C., Jenner, T., Davis, E., and O’Neill, P. (2004) Processing of clustered DNA damage generates additional DSB in mammalian cells post-irradiation. Nucleic Acids Res. 32, 1602−1609. (23) Malyarchuk, S., Castore, R., and Harrison, L. (2008) DNA repair of clustered lesions in mammalian cells: involvement of non homologous end-joining. Nucleic Acids Res. 36, 4872−4882. (24) Kozmin, S. G., Sedletska, Y., Reynaud-Angelin, A., Gasparutto, D., and Sage, E. (2009) The formation of double-strand breaks at multiply damaged sites is driven by the kinetics of excision/incision at base damage in eukaryotic cells. Nucleic Acids Res. 37, 1767−1777. (25) Fujimoto, H., Pinak, M., Nemoto, T., O’Neill, P., Kume, E., Saito, K., and Maekawa, H. (2005) Molecular Dynamics simulation of clustered DNA damage sites containing 8-oxoguanine and abasic site. J. Comput. Chem. 26, 788−798. (26) Hazel, R. D., Tian, K., and de los Santos, C. (2008) NMR Solution Structures of Bistranded Abasic Site lesions in DNA. Biochemistry 47, 11909−11919. (27) Lin, Z., and de los Santos, C. (2001) NMR characterization of clustered bistrand abasic site lesions: effect of orientation on their solution structure. J. Mol. Biol. 308, 341−352. (28) Zálešaḱ , J., Lourdin, M., Krejčι ́, J., Constant, J.-F., and Jourdan, M. (2014) Structure and Dynamics of DNA Duplexes Containing a Cluster of Mutagenic 8-Oxoguanine and Abasic Site Lesions. J. Mol. Biol. 426, 1524−1538. (29) Kappen, L. S., and Goldberg, I. H. (1989) Identification of 2deoxyribonolactone at the site of neocarzinostatin-induced cytosine release in the sequence d(AGC). Biochemistry 28, 1027−1032. (30) Pitié, M., Bernadou, J., and Meunier, B. (1995) Oxidation at Carbon-1′ of DNA Deoxyriboses by the Mn-TMPyP/KHSO5 System Results from a Cytochrome P-450-type Hydroxylation Reaction. J. Am. Chem. Soc. 117, 2935−2936. (31) Pratviel, G., Bernadou, J., and Meunier, B. (1998) DNA and RNA Cleavage by Metal Complexes. Adv. Inorg. Chem. 45, 251−312. (32) Decarroz, C., Wagner, J. R., and Cadet, J. (1987) Specific deprotonation reactions of the pyrimidine radical cation resulting from the menadione mediated photosensitization of 2′-deoxycytidine. Free Radical Res. Commun. 2, 295−301. (33) Von Sonntag, C., and Schuchmann, H. P. (1991) The Elucidation of Peroxyl Radical Reactions in Aqueous Solution with 3906

DOI: 10.1021/acs.biochem.6b00396 Biochemistry 2016, 55, 3899−3906