Normal and reverse base pairing of Iz and Oz lesions in DNA: structural implications for mutagenesis

Abstract

Although G:C to C:G mutations are known to be crucial in some key tumor genes, the exact mechanisms of these mutations are not known. Herein, we have used a density functional theoretical approach to investigate the mechanisms of G:C to C:G mutations induced by 2,5-diamino-4H-imidazol-4-one (Iz), and its hydrolysed product 2,2,4-triamino-5(2H)-oxazolone (Oz) formed due to oxidative degradation of guanine (G) and 8-oxoguanine (8-oxoG). Structures, binding energies and electronic properties of different base pair complexes involving the most stable tautomers of Iz and Oz were studied in detail by considering the nucleobase and nucleoside models. It is shown that the most stable keto tautomer of anti-Iz makes the most stable base pair complex with G by adopting a reverse base pairing orientation. It is further found that the most stable keto tautomer of anti-Oz favors insertion of G opposite it and the resulting base pair complex is stabilized by both the normal and reverse base pairing orientations. As the reverse base pairing structures would be difficult to occur in physiological DNA, it may be proposed that the insertion of G opposite normal Oz may be mainly responsible for the observed G to C mutations in DNA. The results obtained in the present study can be helpful for further structural studies on Iz and Oz lesions in DNA in presence of different DNA polymerases to unravel the mechanisms of G to C mutations observed in carcinogenic genes.

---

1. Introduction

DNA damage products formed due to reactions of in vivo reactive species, photosensitizers, ionizing radiation and oxidizing agents with DNA bases have been found to be interlinked with mutations,^1,2 cancer and several other diseases.^3–5 Among the different DNA bases, guanine (G) is highly susceptible to oxidative damage⁶ that can produce a number of damaged products such as 8-oxoguanine (8-oxoG),^7–12 2,6-diamino-4-oxo-5-formamidopyrimidine (FapyG),^11–15 2,5-diamino-4H-imidazol-4-one (Iz),^16–19 2,2,4-triamino-5(2H)-oxazolone (Oz),^20–25 spiroiminodihydantoin (Sp),^26–30 guanidinihydantoin (Gh),^26–32 iminoallointoin (Ia),³³ etc. The typical mutations induced by oxidatively damaged products of guanine are G:C to T:A and G:C to C:G, which have been observed in some key tumor genes such as p53 and K-ras oncogenes.^34–36 Among the different oxidatively damaged products of guanine, 8-oxoG is the most ubiquitous and often induces G:C to T:A mutations.^37,38 Similarly, FapyG is also believed to induce G:C to T:A mutations in DNA.^39–42 However, the mechanisms of inductions of G:C to C:G mutations are not properly understood and are in continuous debate. As 8-oxoG can be readily degraded to further stable reaction products, it may be proposed that the oxidation products of 8-oxoG may be involved in inducing G:C to C:G mutations.^43–50 As Iz, Oz, Gh, Sp etc. can be formed from the oxidation of guanine and 8-oxoG^51–54 (Scheme 1), and some of these lesions are observed in the liver and colon of mice,^55,56 it is important to rationalize the cellular consequences of these lesions in order to understand mutagenesis and carcinogenesis in detail.

Scheme 1 Formations of Iz, Oz, Gh and Ia lesions due to the oxidations of both G and 8-oxoG. The atomic numbering scheme adopted here corresponds to that used for G.

In an earlier in vivo⁵⁷ study, by using wild-type AB1157 Escherichia coli cells, it was found that Iz is 91% mutagenic and induces primarily G:C to C:G mutations. In another in vitro study, insertion of G opposite Iz was also observed.⁵⁸ Although mutagenic behavior of Iz in bacterial cells is somewhat understood, its mutagenic behavior in eukaryotic cells is even further poorly understood. Further, as Iz can be hydrolyzed to Oz under physiological conditions (half-life of this reaction is 147 min),²² it is also important to understand the Oz-induced mutagenicity in detail. Recently, quantifications of 8-oxoG and Oz at the CpG site in the p53 gene have revealed that formation of Oz gets enhanced appreciably in the presence of 5-methylcytosine.⁵⁹ Similarly, in samples of liver DNA, formation of two to six molecules of Oz per 10⁷ guanine bases have been detected.²⁵ These results indicate that the formations of Iz and Oz may have serious implications in mutagenesis and carcinogenesis.

In order to understand the mutagenic behaviour of Oz, several in vitro nucleotide insertion studies have been undertaken in presence of eukaryotic DNA polymerases.^60–64 For example, by using eukaryotic DNA polymerases (Pols) α, β, ε, and δ, it was found that G is favorably inserted opposite Oz,^60,61 while the use of Klenow Fragment exo- (KF exo-), Pols Taq, γ, IV and ζ resulted in the insertions of both G and adenine (A) opposite Oz.^60–62 Interestingly, nucleotide insertion by Pol η resulted in incorporation of G, A, and cytosine (C) opposite Oz.⁶³ However, primer extension catalysed by Pols η, β, and KF exo- beyond the lesion site was found to be more pronounced for Oz:G base pair compared to those of the Oz:C, Oz:A and Oz:T complexes.⁶⁴ Interestingly, in the case of Pol REV1, the incorporation of C opposite Oz was found to be dominant.⁶¹ Even, this incorporation was found to be more efficient than opposite 8-oxoG and other oxidative lesions.⁶¹ Surprisingly, in the presence of bacterial DNA polymerase, Oz was observed to induce G to T mutation at a frequency an order of magnitude higher than that of 8-oxoG.⁶⁵

In order to understand the stabilities of Iz:G and Oz:G base pairs, interactions of Iz and Oz with G were studied by using ab initio methods without considering DNA backbone.^58,61,66,67 It was found that Iz and Oz bind with G in the inverted orientations (Scheme 2a and b).⁶⁸ In this orientation, Iz undergoes 180 deg. rotation about the N1(Iz)–H1(G) hydrogen bond and makes a reverse Watson–Crick base pair. Similarly, in the inverted orientation, Oz undergoes rotations about the and O51(Oz)–H2(G) and O52(Oz)–H1(G) hydrogen bonds (Scheme 1) to make a reverse Wobble base pair. Although reverse base pair complexes can be synthesized chemically, its occurrence in physiological DNA is rare. Hence, it is necessary to understand if the proposed binding modes of Iz and Oz are indeed the most stable and can occur in duplex DNA. Further, as incorporations of A and C opposite Oz by different eukaryotic DNA polymerases were observed, the mechanisms of these insertions should also be unravelled. Further, it is necessary to understand under which circumstances bacterial polymerases induce G to T mutations for Oz. Additionally, as Iz is the direct oxidation product of G and 8-oxoG and is very slowly degraded to Oz,^22,69 it is also desirable to study whether incorporations of A, C and T are possible opposite Iz. Similarly, as tautomerization of DNA bases is an important phenomenon in biology and can significantly influence base pairing abilities of normal and damaged products of DNA,^39,41,70,71 it is desirable to understand tautomerization effects of Iz and Oz on DNA to completely unravel mechanisms of Iz- and Oz-induced mutagenicity.

Scheme 2 Structures of (a) proposed Iz:G complex,⁵⁸ (b) proposed Oz:G complex,⁶⁰ (c) normal Watson–Crick C:G complex, and (d) reverse Watson–Crick C:G complex. Here dR refers to the N-glycosidic bond.

Owing to the above facts, we have studied here structures, binding energies and electronic properties of different base pair complexes involving the most stable tautomers of Iz and Oz by employing the density functional theory (DFT). DFT studies have been used extensively to unravel structures and energetics of different DNA base lesions.^72–75 As oxidation of G weakens its N-glycosidic bond,⁷⁶ thereby reducing the rotational barrier energy required for conversion of anti-conformation to syn-conformation, different base pair interactions were also studied by considering the anti- and syn-conformations of Iz and Oz. In addition to these, we have also studied stabilities of different base pair complexes involving Iz and Oz in normal and reverse base pairing orientations to completely rationalize mechanisms of mutations induced by these lesions.

2. Computational methodology

2.1 Nucleobase model

Structures and energies of different tautomers of Iz and Oz nucleobases were obtained by using the B3LYP hybrid DFT method^77,78 and 6-31+G* basis set. Subsequently, different base pair complexes involving the most stable tautomers of Iz and Oz nucleobases in the anti- and syn-conformations were optimized at the B3LYP/6-31+G* level of theory. As in the nucleobase model, the deoxyribose group was replaced by the H9 atom, different base pair interactions in the anti- and syn-conformations were studied by considering the Watson–Crick and Hoogstein faces of Iz and Oz respectively. In order to consider the effects of dispersion and electron correlation on total energies of all optimized base pair complexes, ωB97XD/AUG-cc-pVDZ level of theory^79,80 was used for single point energy calculations. All geometry optimization calculations were performed in the aqueous medium by using the integral equation formalism of the polarizable continuum model (IEF-PCM) of the self-consistent reaction field theory (SCRF).^81,82 As geometry optimizations and vibrational analyses were not performed at the ωB97XD/AUG-cc-pVDZ level of theory, zero-point energy (ZPE) corrections obtained at the B3LYP/6-31+G* level of theory were considered to be valid at the former level also. In order to account for dispersion as well as electron correlation effects on the optimized geometries and total energies, the most stable tautomers of Iz and Oz and certain important base pair complexes were further optimized at the B3LYP-D3/AUG-cc-pVDZ level of theories.⁸³ As B3LYP-D3 and ωB97XD methods produce reliable structural and energetic data, the results obtained by these methods will only be discussed here.

2.2 Nucleoside model

In order to ensure that the most stable tautomers of Iz and Oz lesions obtained by considering the nucleobase model would occur in DNA, these tautomers were further optimized by considering the nucleoside model. This helped us to identify the exact conformations of the most stable tautomers of dIz and dOz in DNA (i.e. the χ values, planarity of bases, etc.). In addition to these, the most stable base pair complexes involving Iz and Oz (i.e. Iz:G, Oz:G complexes) in the normal and reverse base pair orientations were also optimized by considering the nucleoside model. All geometry optimizations in the nucleoside model were carried out using the B3LYP/6-31+G* level of theory.

2.3 Binding energy calculation

To calculate the ZPE-corrected binding energies (BE) of the different base-pair complexes, eqn (1) given below was used:

AB_BE = AB_TE − [A_TE + B_TE] (1)

where A and B are any two DNA bases and AB is a base pair complex between A and B. The subscripts BE and TE stand for ZPE-corrected binding energy and ZPE-corrected total energy, respectively. All the calculations were performed using the Gaussian09 suite of program (G09)⁸⁴ and structures were visualized by employing the GaussView program (version 5.0).⁸⁵

3. Results and discussion

3.1 Structures and energies of different tautomers of Iz and Oz

Geometry optimizations at the B3LYP/6-31+G* level of theory yielded twelve tautomeric conformations for Iz, the structures and relative energies of which are presented in Fig. S1 (ESI†). Among these tautomers, the N1-deprotonated keto tautomeric form of Iz is found to be the most stable. Similarly, among eleven possible tautomers of Oz, the N1-deprotonated keto tautomer is the most stable (Fig. S2†). It is further noticed that while the most stable tautomer of Iz has a closed ring structure, the most stable tautomer of Oz has an open ring structure. Interestingly, geometry optimizations yielded two different conformations of the most stable tautomer of Oz, where one tautomer is planar and the other is non-planar. In order to confirm these findings, the most stable tautomers of Iz and Oz were further optimized by using the B3LYP-D3/AUG-cc-pVDZ level of theory. The results obtained at the B3LYP-D3 method are found to be similar to those obtained at the B3LYP method. The structures and relative energies of these tautomers are illustrated in Fig. 1. From this figure, it is clear that the planar structure of Oz is slightly more stable than its non-planar structure. It is also revealed that as Iz and Oz have only one ring each, they do not adopt any rotameric conformation or stereoisomeric configuration like those of Gh,⁴⁹ Sp,⁸⁶ and Ia⁵⁰ lesions (Fig. 1).

Fig. 1 Optimized structures of the most stable tautomers of Iz and Oz obtained by considering the nucleobase (left side) and nucleoside models (right side). The relative total energies of different conformations of the most stable tautomer of Oz obtained at the B3LYP (brackets) and B3LYP-D3 (curly brackets) methods are shown for comparison of stabilities.

In order to obtain structural and energetic data about these tautomers in single-stranded DNA, the most stable tautomers of Iz and Oz were reoptimized at the B3LYP/6-31+G* level of theory by considering their nucleoside models. The results obtained are depicted in Fig. 1. The important geometrical parameters are provided in Fig. S3.† It was found that the χ values (O4′–C1′–N9–C4 dihedral angles) of dIz and dOz lie within ∼91–101 deg. It should be mentioned that the typical χ values of the anti-conformation of 2′-deoxyguanosine lie in the range −120 to −180 deg., while in the syn-conformation, these lie in the range 0 to 90 deg.⁸⁷ This implies that Iz and Oz may adopt a high syn-conformation in single-stranded DNA (Fig. S3†). This is not surprising as Gh was observed to adopt a high syn-conformation in duplex DNA.^88,89 In addition to this, it is further found that the non-planar tautomer of dOz is about 0.61 kcal mol⁻¹ more stable than the planar tautomer of dOz (Fig. 1). This difference in stability may increase in the duplex DNA due to the ability of non-planar dOz to make intra-nucleotide interactions.

3.2 Structures and binding energies of different base pair complexes involving Iz

Interactions of G with both the anti-Iz and syn-Iz yielded four different base pairs, two each in normal and reverse base pairing orientations. The structures of these complexes are illustrated in Fig. S4.† The ZPE-corrected binding energies of these complexes are presented in Table S1.† From this figure, it is evident that in the normal base pairing orientations, both anti-Iz and syn-Iz are paired with G by two hydrogen bonds, while in the reverse orientations, anti-Iz:G and syn-Iz:G are stabilized by three and two hydrogen bonds respectively. All the complexes are found to be planar and hence may not affect base pairing and stacking interactions with the neighbouring nucleotides. If we compare binding energies of these complexes, it is evident that reverse pairing between anti-Iz and G produces the most stable base pair complex (Fig. 2c), which is about 5 kcal mol⁻¹ more stable than the anti-Iz:G normal complex (Fig. 2a) (Table S1†). It is further revealed that in the reverse orientation, the Watson–Crick face of Iz resembles that of C. However, as the N9 atom is not a part of the 5-membered ring of Iz, the O4(Iz)–H2(G) hydrogen bond is somewhat stretched compared to the corresponding O2(C)–H2(G) hydrogen bond in the normal C:G complex. As a result, the anti-Iz:G reverse base pair complex becomes less stable than the C:G complex by ∼2 kcal mol⁻¹ (Table S1†). However, this complex is ∼3 kcal mol⁻¹ more stable than the mutated T:G complex (Table S1†). This indicates that the insertion of G opposite Iz stabilized by the reverse base pairing interactions would lead to the formation of a stable Iz:G complex as proposed earlier.⁶⁰

Fig. 2 Optimized structures of (a and b) normal and (c and d) reverse anti-Iz:G base pair complexes obtained by considering the nucleobase (left side, b and c) and nucleoside (right side b and d) models. The ZPE-corrected binding energies (kcal mol⁻¹) obtained using the B3LYP and B3LYP-D3 (parentheses) methods, a few important geometrical parameters and 3′- and 5′-terminals are shown for comparison.

To test if the reverse base pair complex can occur in DNA and to find out the conformational details of this complex, dIz:dG normal and reverse base pair complexes were reoptimized by considering the nucleoside models. The optimized structures of these complexes are presented in Fig. 2. As can be found from these figures, the binding of dG with normal and inverted dIz produces two different complexes, where the C1′–C1′ distances are found to be larger (by ∼2.0 Å) than the typical C1′–C1′ distances observed for C:G base pairs in B DNA. However, the χ values of dIz and dG in these complexes are significantly different than those of the C:G complex in B DNA. These results imply that the conversion of dG to dIz will appreciably affect the structure of DNA locally. It should be mentioned that although the C1′–C1′ distance and χ values of the normal and inverted dIz:dG complexes are not very much different, in the latter complex, the 3′- and 5′-sugar-phosphate bonds undergo significant rotation to adopt the reverse-base pair alignment, which is difficult to accomplish in physiological DNA. This is consistent with the facts that in physiological conditions, purines and pyrimidines bases pair with the complementary bases of DNA by adopting only the normal Watson–Crick orientations (Scheme 2c).

Among the different normal and reverse base pair complexes formed due to interaction of C with Iz (Fig. S5†), the anti-Iz:C complex is found to be the most stable. In this complex, C is bonded with Iz in the normal base pairing orientation by making two hydrogen bonds. As the normal G:C complex is stabilized by three hydrogen bonds, it is about 5 kcal mol⁻¹ more stable than the anti-Iz:C normal complex. This indicates that the conversion of G to Iz will significantly diminish its binding strength with C. It is further found that the anti-Iz:C normal complex is ∼3 kcal mol⁻¹ less stable than the anti-Iz:G reverse complex (Table S1†). Hence, the incorporation of C opposite Iz in DNA would be less preferred than that of G.

Among the different Iz:A complexes (Fig. S6†), the anti-Iz:A complex paired in the normal base pairing orientation is the most stable. However, this complex is ∼5 kcal mol⁻¹ less stable than the anti-Iz:G reverse base pair complex and ∼1 kcal mol⁻¹ less stable than the anti-Iz:C normal base pair complex (Table S1†). This implies that the incorporation of A opposite Iz would not be preferred in DNA.

Interestingly, stabilities of anti-Iz:T normal and reverse complexes are found to be similar (Fig. S7 and Table S1†). However, these complexes are ∼4 kcal mol⁻¹ less stable than the anti-Iz:G reverse complex. Hence the incorporation of T opposite Iz would not be favoured in DNA.

If we compare stabilities of all the base pair complexes involving Iz, the following order is found: anti-Iz:G (reverse paring) > anti-Iz:C (normal pairing) > Iz:T (normal/reverse pairing) > Iz:A (normal) > Iz:G (normal) (Table S1†). This shows that the normal base pairing between Iz and G is the least stable and the reverse base pairing between the same nucleobases is the most stable. However, occurrence of Iz:G reverse base pair complex in normal physiological DNA is difficult and will substantially affect neighboring nucleotides. Hence, the predominant insertion of G opposite Iz observed in earlier biochemical studies may be understood in the following ways. (1) It may possible that the insertion of G opposite Iz followed by normal base pairing is further stabilized by active site aminoacid residues of replicative DNA polymerases. Similar results were obtained for Gh, Sp and other DNA lesions, where a particular conformation was found to be stabilized by aminoacid residues of DNA polymerases.^89–92 If this happens, then the least stable Iz:G complex may become highly stable in the presence of DNA polymerases. (2) As DNA polymerases are tightly packed, rotation of the sugar-phosphate group of Iz to adopt the reverse orientation may not be possible inside the active site of DNA polymerases. Hence, Iz may undergo rotation to adopt the reverse orientation in DNA prior to the insertion of G by DNA polymerases. In this situation, Iz:G reverse base-pairing structure may occur in DNA. However, these presumptions need to be verified by performing high resolution structural studies and are beyond the scope of the present study.

3.3 Structures and binding energies of different base pair complexes involving Oz

As Oz has an open ring structure, it is more flexible than Iz and hence can base pair with the different bases of DNA in several possible ways. This is quite evident from Fig. S8–S11,† where Oz is found to make base pair interactions with G, C, A, and T in nine, eight, eight, and five different ways respectively. Among the different normal Oz:G base pair complexes, the anti-Oz1:G_normal complex is the most stable (Fig. 3a), where Oz1 is non-planar and binds with G by making two strong hydrogen bonds by employing its two O5 atoms. Similarly, among the different Oz:G reverse base pair complexes, the non-planar anti-Oz1:G_reverse complex is the most stable (Fig. 3c). In this complex, Oz is also making two hydrogen bonds with G by engaging its O5 atoms in a manner similar to that of the anti-Oz1:G_normal complex. As the binding patterns of these complexes are similar, these are found to be isoenergetic (Table S2,† Fig. 3a and c).

Fig. 3 The optimized structures of different anti-Oz:G non-planar (a–d) and planar (e–h) complexes that possess (a, b, e and f) normal and (c, d, g and h) reverse base pairing alignments obtained by using nucleobase (left side) and nucleoside models (right side). The binding energies (kcal mol⁻¹) obtained using the B3LYP and B3LYP-D3 (parentheses) methods, a few important geometrical parameters and 3′- and 5′-terminals are shown for comparison.

Notably, geometry optimizations yielded two more isoenergetic Oz:G base pair complexes (i.e. anti-Oz2:G_normal and anti-Oz2:G_reverse), where Oz is found to be planar (Fig. 3e and g). It should be noted that in an earlier study, the planar anti-Oz2:G_reverse base pair complex (Fig. 3g and Scheme 2b) was found to be the most stable and was presumed to be interlinked with G to C mutations in DNA.⁶⁰ However, herein, this complex is found to be about 1 kcal mol⁻¹ less stable than the planar normal (anti-Oz1:G_normal) and reverse (anti-Oz1:G_reverse) base pair complexes (Table S2† and Fig. 3). It should be mentioned that in that study,⁶⁰ authors had optimized geometries of different Oz:G complexes extracted from different minimization stages of a molecular mechanics (MM) study. As minimization algorithms of a MM-simulation study is less accurate than that of a DFT study, the entire conformational spaces of Oz:G base pair complex may not have been properly sampled in the MM-minimization study. This may be the reason, why the authors were unable to identify the most stable Oz:G complex as obtained here (Fig. 3a and b).

In order to ensure that the most stable non-planar complexes would remain the most stable in double-stranded DNA also, all planar and non-planar complexes of Oz:G were re-optimized at the B3LYP/6-31+G* level of theory by considering their nucleoside models. The structures and binding energies of these complexes are depicted in Fig. 3. From this figure, it is clear that the non-planar anti-dOz1:dG_normal complex is about 1 kcal mol⁻¹ more stable than the non-planar anti-dOz1:dG_reverse base pair complex. Similarly, the planar anti-dOz2:dG_normal base pair complex is about 1 kcal mol⁻¹ more stable than the corresponding reverse base pair complex. This shows that the reverse base pair complexes would be less stable in DNA than those of the normal base pair complexes. This is in agreement with the results obtained by using the nucleobase models. Further, as in the non-planar anti-Oz:G_normal complex, two amino groups are exposed to environment, these may be involved in additional interactions with neighboring nucleotides, thereby further enhancing its stability.^93,94 Further, as replicative DNA polymerases insert nucleotides opposite DNA base lesions mainly based on steric interactions and ability of the base lesion to make hydrogen bonds with different active site amino acid residues, insertion of G opposite non-planar Oz may become more favorable in DNA. This is because, the non-planar Oz has a few free hydrogen bond acceptor and donor sites, which may be involved in making hydrogen bonds with active site residues of replicative DNA polymerases to stabilize the non-planar anti-Oz1:G_normal complex. Moreover, due to the reasons explained for the Iz:G reverse base pair complex, occurrence of Oz:G reverse base pair complex may not be feasible in DNA.

Among different Oz:C complexes (Fig. S9†), the non-planar anti-Oz1:C complex was found to be the most stable. In this complex, Oz binds with C by adopting normal base pairing orientation and by making three hydrogen bonds like those of the normal G:C base pair complex (Fig. 4). Further, the distributions of electrostatic potentials and electron densities around these complexes are found to be similar (Fig. 4). It should be noted that in spite of structural likelihood of anti-Oz1:C and G:C normal complexes, the former complex is about 4 kcal mol⁻¹ less stable than the latter complex (Tables S1, S2† and Fig. 4). This indicates that the conversion of G to Oz may affect stability of DNA near the lesion site. Interestingly, binding energies of anti-Oz1:G and anti-Oz1:C normal non-planar complexes are found to be similar. This indicates that insertion of G and C opposite the non-planar Oz may produce equally stable complexes in the duplex DNA. This may be the reason why translesion synthesis polymerases like Pol η that mainly insert nucleotides opposite a base lesion on the basis of base pair alignments, hydrogen bonding interactions and electrostatic potentials, incorporate C opposite Oz in DNA. In this scenario, Oz will be non-mutagenic. Further, as while making base pair interactions with C, Oz utilizes all of its hydrogen bond donor and acceptor groups, it may not efficiently interact with residues of replicative DNA polymerases. Presumably for this reason, insertion of C facilitated by replicative DNA polymerases opposite Oz was not observed in DNA.

Fig. 4 Optimized structures and electron densities mapped onto electrostatic potentials of (a) G:C and (b) anti-Oz1:C base pair complexes obtained at the B3LYP-D3/AUG-cc-pVDZ level of theory. The binding energies (in brackets) (kcal mol⁻¹) of these complexes obtained at the above level of theory are also shown for comparison of stabilities.

In the case of Oz:A complexes (Fig. S10†), it is found that the hydrogen bonds made by O5 and two amino groups of non-planar anti-Oz1 with A produce the most stable Oz:A complex (Fig. 5b). In this complex, Oz has also adopted a normal base pairing orientation. However, this complex is about 2 kcal mol⁻¹ less stable than the most stable anti-Oz1:G normal non-planar complex. Interestingly, the stability of this complex is comparable to that of the Watson–Crick T:A complex (Fig. 5a), as both the complexes are stabilized by two hydrogen bonds each. Further, the binding patterns and electronic environments of these complexes are found to be similar (Fig. 5). This may be the reason of incorporation of A opposite Oz by translesion synthesis polymerases and A-family polymerases that are involved in replication of mitochondrial genome.⁶⁰

Fig. 5 Optimized structures and electron densities mapped onto the electrostatic potentials of (a) T:A and (b) the most stable anti-Oz1:A complexes obtained at the B3LYP-D3/AUG-cc-pVDZ level of theory. The binding energies (in brackets) (kcal mol⁻¹) of these complexes are shown for comparison of stabilities.

Among different possible Oz:T complexes (Fig. S11†), the anti-Oz1:T non-planar complex bound in the normal base pair orientation is the most stable (Fig. S11a†). However, this complex is about 3 kcal mol⁻¹ less stable than the most stable anti-Oz1:G non-planar normal complex (Fig. 3). If we compare stabilities of different complexes involving Oz, the following order is found: Oz:C ≥ Oz:G > Oz:A > Oz:T. This clearly indicates that the insertion of T opposite Oz is the least favoured. This may be the reason, why DNA polymerases do not insert T opposite Oz. These results are in agreement with the earlier primer extension and thermal denaturation studies.⁶⁴

A comparison of different base pair complexes involving Iz and Oz reveals that Iz makes only planar complexes and Oz can make both the planar and non-planar complexes. Further, the non-planar complexes of Oz are the most stable. It is further found that the insertions of G opposite both Iz and Oz in the anti-conformations are highly preferred. However, while Iz:G reverse base pair structure is the most stable, the Oz:G complex is stabilized by both normal and reverse base pairing orientations. Further, if we compare binding energies of Gh:G (−11.00 kcal mol⁻¹), Ia:G (−11.00 kcal mol⁻¹), Iz:G (−9.48 kcal mol⁻¹) and Oz:G (−12.20 kcal mol⁻¹) normal base pair complexes,^49,50 it is clear that the last complex is the most stable. Similarly, if we compare the stabilities of the above base pair complexes stabilized in the reverse base pair orientations, the following order is found: Ia:G (−15.20) > Iz:G (−14.10) > Oz:G (−12.22).^49,50 This indicates that the insertion of G opposite reverse Ia is more stable than those of Iz and Oz. However, as the reverse base-pairing structure is difficult to accomplish in physiological DNA, insertion of G opposite normal Oz may be mainly responsible for the observed G to C mutations in DNA as observed in various tumor genes. However, consideration of base stacking, sequence effects and active sites of DNA polymerases may provide further insights.

4. Conclusion

It is found that the most stable keto tautomer of Iz has a closed ring planar structure, while the most stable keto tautomer of Oz has an open ring structure. Although the most stable tautomer of Oz can adopt both the planar and non-planar conformations, insertion of complementary bases opposite non-planar Oz would be preferred. Among the different base pair complexes involving Iz, the Iz:G complex is the most stable. This complex is mainly stabilized by reverse base pairing interactions between the anti-conformation of Iz and G. Similarly, among the different base pair complexes involving Oz, the anti-Oz:G complex is the most stable. Interestingly, it is found that this complex can be formed by both the normal and reverse pairing between anti-Oz and G. Although the pairing of G with inverted Iz provides more stability than its pairing with normal Oz, due to the reverse base pairing nature of the resulting complex, occurrence of Iz:G in DNA may not be feasible. On the basis of these results, it may be proposed that the insertion of G opposite normal Oz in DNA may be responsible for the observed G to C mutations in DNA.

Acknowledgements

NRJ is thankful to IIITDM Jabalpur for a research initiation grant. PCM is thankful to the National Academy of Sciences, India (NASI) for a Senior Scientist Fellowship.

References

1. D. Wang, D. A. Kreutzer and J. M. Essigman, Mutat. Res., 1998, 400, 99.
2. K. C. Cheng, D. S. Cahill, H. Kasai, S. Nishimura and L. A. Loeb, J. Biol. Chem., 1992, 267, 166.
3. M. S. Cooke, M. D. Evans, M. Dizdaroglu and J. Lunec, FASEB J., 2003, 17, 1195.
4. W. L. Neeley and J. M. Essigmann, Chem. Res. Toxicol., 2006, 19, 491.
5. S. Kawanishi, Y. Hiraku and S. Oikawa, Mutat. Res., 2001, 488, 65.
6. S. Steenken and S. V. Jovanovic, J. Am. Chem. Soc., 1997, 119, 617.
7. N. R. Jena and P. C. Mishra, J. Phys. Chem. B, 2005, 109, 14205.
8. H. J. Helbock, K. B. Beckman, M. K. Shigenaga, P. B. Walter, A. A. Woodall, H. C. Yeo and B. N. Ames, Proc. Natl. Acad. Sci. U. S. A., 1998, 95, 288.
9. J.-L. Ravanat, T. Douki, P. Duez, E. Gremaud, K. Herbert, T. Hofer, L. Lasserre, C. Saint-Pierre, A. Favier and J. Cadet, Carcinogenesis, 2002, 23, 1911.
10. J.-L. Ravanat, C. Saint-Pierre and J. Cadet, J. Am. Chem. Soc., 2003, 125, 2030.
11. J. Cadet, T. Douki and J.-L. Ravanat, Acc. Chem. Res., 2008, 41, 1075.
12. N. R. Jena and P. C. Mishra, Free Radical Biol. Med., 2012, 53, 81.
13. B. Tudek, J. Biochem. Mol. Biol., 2003, 36, 12.
14. J. P. Pouget, S. Frelon, J.-L. Ravanat, I. Testard, F. Odin and J. Cadet, Radiat. Res., 2002, 157, 589.
15. T. Douki, R. Martini, J.-L. Ravanat, R. J. Turesky and J. Cadet, Carcinogenesis, 1997, 18, 2385.
16. C. Vialas, G. Pratviel, C. Claparols and B. Meunie, J. Am. Chem. Soc., 1998, 120, 11548.
17. C. Vialas, C. Claparols, G. Pratviel and B. Meunie, J. Am. Chem. Soc., 2000, 122, 2157.
18. N. R. Jena, J. Biosci., 2012, 37, 503.
19. K. Kino and I. Saito, J. Am. Chem. Soc., 1988, 120, 7373–7374.
20. J. L. Ravanat and J. Cadet, Chem. Res. Toxicol., 1995, 8, 379.
21. D. Gasparutto, J. Ravanat, O. Gerot and J. Cadet, J. Am. Chem. Soc., 1998, 120, 10283.
22. J. Cadet, M. Berger, G. W. Buchko, P. C. Joshi, S. Raoul and J.-L. Ravanat, J. Am. Chem. Soc., 1997, 116, 7403.
23. T. Douki and J. Cadet, Free Radical Res., 1996, 24, 369.
24. T. Douki and J. Cadet, Int. J. Radiat. Biol., 1999, 75, 571.
25. B. Matter, D. Malejka-Giganti, A. S. Csallany and N. Tretyakova, Nucleic Acids Res., 2006, 34, 5449.
26. C. J. Burrows, J. G. Muller, O. Kornyushyna, W. Luo, V. Duarte, M. D. Leipold and S. S. David, Environ. Health Perspect., 2002, 110, 713.
27. W. Luo, J. G. Muller, E. M. Rachlin and C. J. Burrows, Org. Lett., 2000, 2, 613.
28. J. C. Niles, J. S. Wishnok and S. R. Tannenbaum, Org. Lett., 2001, 3, 963.
29. A. M. Fleming, J. G. Muller, A. C. Dlouhy and C. J. Burrows, J. Am. Chem. Soc., 2012, 134, 15091.
30. L. Jia, V. Shafirovich, R. Shapiro, N. E. Geacintov and S. Broyde, Biochemistry, 2005, 44, 6043.
31. Y. Ye, B. H. Munk, J. G. Muller, A. Cogbill, C. J. Burrows and H. B. Schlegel, Chem. Res. Toxicol., 2009, 22, 526.
32. Y. Ye, J. G. Muller, W. Luo, C. L. Mayne, A. J. Shallop, R. A. Jones and C. J. Burrows, J. Am. Chem. Soc., 2003, 125, 13926.
33. Z. Zhu, A. M. Fleming, A. M. Orendt and C. J. Burrows, J. Org. Chem., 2016, 81, 351.
34. J. Dorszewska and Z. Adamczewska-Gonerzewicz, Respir. Physiol. Neurobiol., 2004, 139, 227.
35. S. Ohnishi, H. Saito, N. Suzuki, N. Ma, Y. Hiraku, M. Murata and S. Kawanishi, Biochem. Biophys. Res. Commun., 2011, 413, 236.
36. K. G. Higinbotham, J. M. Rice, B. A. Diwan, K. S. Kasprzak, C. D. Reed and A. O. Perantoni, Cancer Res., 1992, 52, 4747.
37. S. Shibutani, M. Takeshita and A. P. Grollman, Nature, 1991, 349, 431.
38. M. Moriya, Proc. Natl. Acad. Sci. U. S. A., 1993, 90, 1122.
39. T. H. Gehrke, U. Lischke, K. L. Gasteiger, S. Schneider, S. Arnold, H. C. Muller, D. S. Stephenson, H. Zipse and T. Carell, Nat. Chem. Biol., 2013, 9, 455.
40. C. J. Wiederholt and M. M. Greenberg, J. Am. Chem. Soc., 2002, 124, 7278.
41. N. R. Jena and P. C. Mishra, ChemPhysChem, 2013, 14, 3263.
42. N. R. Jena, A. E. Mark and P. C. Mishra, ChemPhysChem, 2014, 15, 1779.
43. K. Kino and H. Sugiyama, Chem. Biol., 2001, 8, 369.
44. O. Kornyushyna, A. M. Berges, J. G. Muller and C. J. Burrows, Biochemistry, 2002, 41, 15304.
45. O. Kornyushyna and C. J. Burrows, Biochemistry, 2003, 42, 13008.
46. V. Duarte, J. G. Muller and C. J. Burrows, Nucleic Acids Res., 1999, 27, 496.
47. P. T. Henderson, J. C. Delaney, J. G. Muller, W. L. Leeley, S. R. Tannenbaum, C. J. Burrows and J. M. Essigmann, Biochemistry, 2003, 42, 9257.
48. M. Suzuki, K. Kino, T. Kawada, M. Morikawa, T. Kobayashi and H. Miyazawa, Chem. Res. Toxicol., 2015, 28, 1307.
49. N. R. Jena, V. Gaur and P. C. Mishra, Phys. Chem. Chem. Phys., 2015, 17, 18111.
50. N. R. Jena, M. Bansal and P. C. Mishra, Phys. Chem. Chem. Phys., 2016, 18, 12774.
51. G. R. Martinez, M. H. G. Medeiros, J. L. Ravanat, J. Cadet and P. Di Mascio, Biol. Chem., 2002, 383, 607.
52. B. H. Munk, C. J. Burrows and H. B. Schlegel, J. Am. Chem. Soc., 2008, 130, 5245.
53. S. S. Hah, H. M. Kim, R. A. Sumbad and P. T. Henderson, Bioorg. Med. Chem. Lett., 2005, 15, 3627.
54. W. Luo, J. G. Muller, E. M. Rachlin and C. J. Burrows, Chem. Res. Toxicol., 2001, 14, 927–938.
55. A. Mangerich, C. G. Knutson, N. M. Parry, S. Muthupalani, W. Ye, E. Prestwich, L. Cui, J. L. McFaline, M. Mobley, Z. Ge, K. Taghizadeh, J. S. Wishnok, G. N. Wogan, J. G. Fox, S. R. Tannenbaum and P. C. Dedon, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, E1820.
56. B. Matter, D. Maleijka-Giganti, A. S. Csallany and N. Tretyakova, Nucleic Acids Res., 2006, 34, 5449.
57. W. L. Neeley, J. C. Delaney, P. T. Henderson and J. M. Essigmann, J. Biol. Chem., 2004, 279, 43568.
58. K. Kino and H. Sugiyama, Chem. Biol., 2001, 8, 369.
59. X. Ming, B. Matter, M. Song, E. Veliath, R. Shanley, R. Jones and N. Tretyakova, J. Am. Chem. Soc., 2014, 136, 4223.
60. K. Kino, K. Sugasawa, T. Mizuno, T. Bando, H. Sugiyama, M. Akita, H. Miyazawa and F. Hanaoka, ChemBioChem, 2009, 10, 2613.
61. M. Suzuki, K. Kino, T. Kawada, M. Morikawa, T. Kobayashi and H. Miyazawa, Chem. Res. Toxicol., 2015, 28, 1307.
62. V. Duarte, D. Gasparutto, M. Jaquinod and J. Cadet, Nucleic Acids Res., 2000, 28, 1555.
63. K. Kino, N. Ito, K. Sugasawa, H. Sugiyama and F. Hanako, Nucleic Acids Symp. Ser., 2004, 48, 171.
64. M. Suzuki, K. Ohtsuki, K. Kino, T. Kobayashi, M. Morikawa, T. Kobayashi and H. Miyazawa, J. Nucleic Acids, 2014, 178350, 1.
65. P. T. Henderson, J. C. Delaney, F. Gu, S. R. Tannenbaum and J. M. Essigman, Biochemistry, 2002, 41, 914.
66. M. Suzuki, K. Kino, M. Morikawa, T. Kobayashi, R. Komori and H. Miyazawa, Molecules, 2012, 17, 6705.
67. M. Suzuki, K. Kino, M. Morikawa, T. Kobayashi and H. Miyazawa, Molecules, 2014, 19, 11030.
68. X.-L. Yang, H. Sugiyama, S. Ikeda, I. Saito and A. H.-J. Wang, Biophys. J., 1998, 75, 1163.
69. L. Cui, W. Ye, E. G. Prestwich, J. S. Wishnok, K. Taghizadeh, P. C. Debon and S. R. Tannenbaum, Chem. Res. Toxicol., 2013, 26, 195.
70. Y. H. Jang, W. A. Goddard, K. T. Noyes, L. C. Sowers, S. Hwang and D. S. Chung, Chem. Res. Toxicol., 2002, 15, 1023.
71. B. T. Psciuk and B. H. Schlegel, J. Phys. Chem. B, 2013, 117, 9518.
72. P. Kathuria, P. Sharma, M. N. Abendong and S. D. Wetmore, Biochemistry, 2015, 54, 2414.
73. N. R. Jena and M. Bansal, Phys. Biol., 2011, 8, 046007.
74. L. Jia, V. Shafirovich, R. Shapiro, N. E. Geacintov and S. Broyde, Biochemistry, 2006, 45, 6644.
75. K. A. Wilson and S. D. Wetmore, Chem. Res. Toxicol., 2014, 27, 1310.
76. K. Kino, M. Takao, H. Miyazawa and F. Hanaoka, Mutat. Res., 2012, 734, 73.
77. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785.
78. B. Miehlich, A. Savin, H. Stoll and H. Preuss, Chem. Phys. Lett., 1989, 157, 200.
79. J.-D. Chai and M. Head-Gordon, Phys. Chem. Chem. Phys., 2008, 10, 6615.
80. J.-D. Chai and M. Head-Gordon, Chem. Phys., 2008, 128, 084106.
81. J. Tomasi, B. Mennucci and R. Cammi, Chem. Rev., 2005, 105, 2999.
82. G. Scalmani and M. J. Frissh, J. Chem. Phys., 2010, 132, 114110.
83. S. Grimme, J. Antony, S. Ehrlich and H. Krieg, J. Chem. Phys., 2010, 132, 154104.
84. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, V. Nakai, T. Vreven, J. Jr. J. A. Montgomery, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian09, Revision A.1, Gaussian, Inc., Wallingford CT, 2009.
85. R. Dennington, T. Keith and J. Millam, GaussView, Version5, Semichem Inc., Shawnee Mission, KS, 2009.
86. S. Ding, L. Jia, A. Durandin, C. Crean, A. Kolbanovskiy, V. Shafirovich, S. Broyde and N. E. Geacintov, Chem. Res. Toxicol., 2009, 6, 1189.
87. B. Karwowski, F. Dupeyrat, M. Bardet, J.-L. Ravanat, P. Krajewski and J. Cadet, Chem. Res. Toxicol., 2006, 19, 1357.
88. P. Aller, Y. Ye, S. S. Wallace, C. J. Burrows and S. Doublie, Biochemistry, 2010, 49, 2502.
89. J. Beckman, M. Wang, G. Blaha, J. Wang and W. H. Konigsberg, Biochemistry, 2010, 49, 8554.
90. B. E. Eckenroth, A. M. Fleming, J. B. Sweasy, C. J. Burrows and S. Doublie, Biochemistry, 2014, 53, 2075.
91. B. D. Freudenthal, W. A. Beard and S. H. Wilson, Nucleic Acids Res., 2013, 41, 1848.
92. O. Rechkoblit, L. Malinina, Y. Cheng, N. E. Geacintov, S. Broyde and D. J. Patel, Structure, 2009, 17, 725.
93. U. Bren, V. Martinek and J. Florian, J. Phys. Chem. B, 2006, 110, 10557.
94. P. A. Fagan, C. Fabrega, R. Eritja, R. M. F. Goodman and D. E. Wemmer, Biochemistry, 1996, 35, 4026.

---

Footnote

† Electronic supplementary information (ESI) available: Structures of different tautomers of Iz and Oz, structures and ZPE-corrected binding energies of different base pair complexes involving Iz and Oz, and cartesian coordinates of important base pair complexes. See DOI: 10.1039/c6ra14031a

---