Quantum Chemical Studies of the Structure and Stability of N

Department of Chemistry and Biochemistry, University of Lethbridge, 4401 University Drive West, Lethbridge T1K 3M4, Alberta, Canada. J. Phys. Chem. ...
0 downloads 8 Views 4MB Size
Download PDF
Article Cite This: J. Phys. Chem. A XXXX, XXX, XXX−XXX

pubs.acs.org/JPCA

Quantum Chemical Studies of the Structure and Stability of N‑Methylated DNA Nucleobase Dimers: Insights into the Mutagenic Base Pairing of Damaged DNA Published as part of The Journal of Physical Chemistry virtual special issue “Manuel Yánẽ z and Otilia Mó Festschrift”. Lindey R. Felske, Stefan A. P. Lenz, and Stacey D. Wetmore* Department of Chemistry and Biochemistry, University of Lethbridge, 4401 University Drive West, Lethbridge T1K 3M4, Alberta, Canada S Supporting Information *

ABSTRACT: DNA is constantly under attack from exogenous and endogenous sources that modify the chemical structure of the nucleobases. A common type of nucleobase damage is N-methylation, which can result in mutagenesis. Nevertheless, these lesions are often repaired by the DNA repair enzyme AlkB, albeit at varying rates. Herein we use density functional theory (B3LYP-D3(BJ)/6-311++G(2df,2p)//B3LYP/6-31G(d,p)) to comprehensively examine the structural and energetic properties of base pairs between seven nucleobase lesions resulting from N-methylation on the Watson−Crick (WC) binding face and each canonical nucleobase. By characterizing 105 stable nucleobase dimers, we provide fundamental details regarding the preferred lesion base pairings. Specifically, we reveal that the flexibility of the methylamino group resulting from methylation of an exocyclic amino substituent allows the 2MeG, 4MeC, and 6MeA lesions to maintain a preference for canonical WC base pairing, which correlates with the experimentally reported lack of mutagenicity for these damage products. In contrast, calculated distortions in key structural parameters and altered binding energies for base pairs involving adducts formed upon methylation of a ring nitrogen (namely, 1MeG, 3MeT, 1MeA, and 3MeC) help rationalize the associated mutagenicity and repair efficiencies. Most importantly, our work provides molecular-level information about the interactions between N-methylated and canonical nucleobases that is critical for future large-scale modeling of damaged DNA and enzyme−DNA complexes that strive to further uncover the mutagenicity and repair propensities of these detrimental lesions.

■

halides, which can also lead to 3MeC.7 Alternatively, nucleobase methylation, including the formation of N6methyladenine (6MeA) and N4-methylcytosine (4MeC), are common postreplicative modifications that regulate replication initiation and gene expression.8,9 Several DNA repair pathways exist that can reverse the potential detrimental effects of nucleobase modifications.10 One enzyme that targets a range of DNA methyl lesions is AlkB of Escherichia coli, which is an Fe(II)/α-ketogluturate-dependent dioxygenase that catalyzes the oxidative repair of methylation, as well as other types of alkylation damage.11 The efficiency of AlkB varies with the damaged nucleobase site.11−17 Specifically, the best substrates for AlkB-mediated repair are cationic 1MeA and 3MeC, and in general A or C damage products are better substrates for AlkB than T or G lesions.12,15,16 The activity of AlkB is further complicated by the ability to repair both single-

INTRODUCTION

The genetic blueprint of an organism is contained within the sequence of nucleobases in DNA. This information is maintained upon cell division, as DNA is replicated due to complementary Watson−Crick (WC) hydrogen bonding between canonical adenine (A) and thymine (T), and guanine (G) and cytosine (C; Figure 1). Unfortunately, the critical information contained within DNA can be altered when one of the nucleobases is damaged through processes such as oxidation,1 deamination,2 or methylation.3 Nucleobase methylation is particularly interesting due to the large number of DNA methylation pathways. For example, DNA nucleobases can be methylated upon exposure to various agents such as endogenous S-adenosylmethionine, which commonly leads to N1-methyladenine (1MeA, Figure 1), or exogeneous agents such as methylmethanesulfonate, which yields N3-methylthymine (3MeT) and N3-methylcytosine (3MeC).4,5 Other methylating sources include nitrosamines from tobacco smoke, which typically result in N1- or N2-methylguanine (1MeG or 2MeG, Figure 1),6 and naturally occurring methyl © XXXX American Chemical Society

Received: October 23, 2017 Revised: November 29, 2017 Published: November 30, 2017 A

DOI: 10.1021/acs.jpca.7b10485 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

consequences. For example, 1MeA is not mutagenic, while 3MeC is ∼30% mutagenic, predominantly leading to C → T and C → A point mutations. Alternatively, neutral 1MeG is ∼80% mutagenic, leading to G → T, G → A, and G → C mutations, while neutral 3MeT is ∼60% mutagenic, resulting in T → A and T → C mutations. Unsurprisingly, the lesions formed upon methylation of the exocyclic amino group (2MeG, 6MeA, and 4MeC) are generally not mutagenic due to the flexibility of the resulting methylamino substituent.11 Several studies have characterized the structure of DNAcontaining methylated lesions paired opposite canonical nucleobases. Specifically, high-resolution solution NMR of damaged DNA duplexes has shown that the Hoogsteen (H) face of 1MeA hydrogen bonds with the WC face of an opposing T.19 Using a cross-linked protein−DNA system, crystal structures of DNA containing 1MeA, 6MeA, or 3MeC have also been obtained.21 Although 6MeA forms WC pairing with T, the H-1MeA:WC-T pair prevails in the crystal structure, and 3MeC does not form interactions with the opposing G. Subsequent data from solution NMR coupled with molecular dynamics (MD) simulations further confirm the H-1MeA:WCT base-pairing geometry.22 Finally, crystal structures of (5′CGCG(4MeC)G)2 indicate that 4MeC:G WC pair does not significantly disrupt Z-DNA duplex structure, with the additional methyl group directed away from the WC face of C.23 In contrast to methylated lesions paired opposite canonical nucleobases, little information is currently available about the structure of methylated base mispairs despite the importance of understanding the base-pairing properties for rationalizing the replication and repair outcomes of damaged DNA. To the best of our knowledge, only one NMR study has examined the base pairing between 1MeA and the four canonical nucleobases.19 This work revealed that the H binding face of 1MeA interacts with T, G, and A, but 1MeA does not form appreciable pairing interactions with C. Although quantum chemical calculations can complement experimental structural data by providing information about potential hydrogen-bonding interactions in nucleobase dimers, nucleobases methylated on the WC face have been understudied. Indeed, only one combined ab initio and density functional theory (DFT) study has considered the 3MeC:G pair and revealed how the formation of 3MeC disrupts the C:G canonical pair.24 Unfortunately, the corresponding 3MeC mispairs and base pairs involving other N-methylated nucleobase lesions remain unexplored. In contrast, ab initio and DFT calculations have been used to examine the effects of methylation of the purines at ring nitrogens removed from the WC binding face (i.e., N3 or N7, Figure 1) or nucleobase exocyclic carbonyl groups (i.e., O6 of G, O2 or O4 of T, and O2 of C, Figure 1) on the structure and stability of the canonical A:T and G:C WC base pairs.24−26 One study examined the stability of O4-methylthymine mispairs with each canonical base and revealed that methylation can influence the base-pairing preference.27 The effects of epigenetic modifications, including 6MeA and 5-methylcytosine (5MeC), have also been considered in models including nucleobase base pairs,28 base pair steps,29 DNA oligomers,30,31 and nucleosome models.32 Collectively, these works have highlighted that methylation at certain DNA sites can significantly alter the structure and stability of canonical WC pairs and higher-order DNA complexes. Since previous computational studies have underscored the importance of examining the impact of DNA nucleobase

Figure 1. Structure of the canonical DNA nucleobases (adenine (A), cytosine (C), guanine (G), and thymine (T)), as well as the Nmethylated lesions considered in the present work. The chemical numbering and the WC and H binding edges are defined for the canonical bases.

stranded (ss) and double-stranded (ds) DNA, with the preference being lesion-dependent.14 Specifically, AlkB is more effective at repairing 1MeA and 3MeC in ssDNA, while the repair of 1MeG and 3MeT is faster in dsDNA. The AlkB repair activity may also be at least partially dictated by the structure and/or stability of the lesion base pair.18,19 Indeed, although 1MeA, 3MeC, and 3MeT are most efficiently processed in base mismatches in dsDNA, AlkB more efficiently targets 1MeG paired opposite complementary C.14 Nine homologues of E. coli AlkB have been identified in humans (ABH1−8 and FTO), with the lesions targeted by ABH2 and ABH3 including all seven damaged bases formed upon addition of a methyl group to the WC hydrogen-bonding face of a canonical nucleobase (Figure 1). Despite methylation at the WC hydrogen-bonding face, the N-methylated nucleobases that contain an exocyclic methylamino group (2MeG, 6MeA, and 4MeC) may maintain WC complementary pairing, while lesions with the methyl group added directly to the nucleobase ring (1MeG, 3MeT, 1MeA, and 3MeC) disrupt WC hydrogen bonding. To understand their biological implications, experimental studies have characterized the mutagenicity of these lesions in AlkB-deficient cells.11,20 Although many of the lesions are strong replication blocks, successful lesion bypass leads to different mutagenic B

DOI: 10.1021/acs.jpca.7b10485 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

the lesion and/or opposing base (Figure 1), which is presented to the complementary strand when the nucleotide adopts the syn orientation (χ = ∠(O4′C1′NXCY) = 0 ± 90°). In the present work, we considered all possible pairing combinations between the WC or H binding face of the methylated lesion and the WC or H binding face of each canonical base, with our nomenclature identifying the binding face involved in each isolated base pair. It is important to emphasize that the present work details gas-phase structures and binding energies for isolated nucleobase pairs. Although these quantities are important descriptors of duplex formation, our results are not directly comparable to experimentally predicted duplex thermodynamic stability, mutagenicity, or repair efficiency, since there are many other factors involved beyond the predicted gas-phase potential energy for dimerization of isolated base pairs (e.g., nucleobase stacking, helix solvation, and steric effects).46 Nevertheless, our calculations provide important structural and energetic information about each lesion base pair that furthers our understanding of nucleic acid chemistry and will direct future large-scale modeling for which fewer pairing combinations can be practically considered.

methylation, the present work uses DFT to comprehensively evaluate the atomic structure and stability of the hydrogenbonded base pairs between N-methylated nucleobases that have an altered WC face and each canonical nucleobase (Figure 1). The nucleobases examined include those formed upon methylation of the exocyclic amino group, which may not affect WC hydrogen bonding (namely, 2MeG, 6MeA, and 4MeC), as well as neutral (1MeG and 3MeT) and cationic (1MeA and 3MeC) lesions formed upon methylation of a ring nitrogen, which directly impacts the WC binding face. Our data reveal the influence of the methyl substituent on the canonical A:T and G:C pairs and uncover the preferred pairing partner(s) for each methylated lesion in isolated dimers. Characterizing the structures and binding energies of the hydrogen-bonded base pairs between methylated and canonical nucleobases is a crucial step toward understanding the reported differential mutagenicity11,20 and repair12,14 of these methylated lesions. It is anticipated that the fundamental information about the discrete interactions involving modified DNA nucleobases obtained from our work will direct future large-scale modeling of damaged DNA helices and the interactions between damaged DNA and critical cellular machinery, including polymerases and repair enzymes.

■

■

RESULTS AND DISCUSSION The suitability of hydrogen-bonded base pairs involving methylated nucleobases within the context of damaged DNA duplexes was evaluated based on the structural and energetic features of the canonical A:T and G:C WC nucleobase pairs calculated at the same level of theory (Figure 2). Specifically,

COMPUTATIONAL DETAILS Optimization and frequency calculations for all monomers and base pairs were performed at the B3LYP/6-31G(d,p) level of theory. B3LYP has been previously successfully used to study the effects of nucleobase methylation on the A:T and G:C base pairs,24,25 as well as the base-pairing properties for a range of other DNA lesions (see, e.g., refs 33−41). Subsequently, the binding energies were determined using B3LYP-D3 with Becke-Johnson (BJ) damping42 and the 6-311++G(2df,2p) basis set. Scaled (0.9857)43 zero-point vibrational corrections, monomer deformation energies, and basis set superposition error corrections calculated according to the Boys and Bernardi scheme44 are included in all reported interaction energies. All calculations were completed using Gaussian 09 (revision D.01).45 Nucleobase models were implemented that replace 2′deoxyribose with a methyl group (dR = CH3, Figure 1), which permits evaluation of key geometric parameters relevant to DNA helices, including the base-pair width (i.e., the R(C1′··· C1′) distance) and the base-pair opening (i.e., the ∠(NXC1′C1′) angle, NX = N1 for pyrimidine lesions and N9 for purine damage). To determine the preferred methyl group conformation in all models, the most stable orientation of each methyl group was first determined for the nucleobase monomers. Subsequently, base pairs were built using the lowest-energy monomer conformations. In the case of methylation of the nucleobase exocyclic amino group, base pairs were considered with the additional methyl substituent directed either away from or toward the WC binding face of the damaged base. Hydrogen-bonded base pairs were explored between each damaged nucleobase and each canonical nucleobase in all possible pairing combinations. Specifically, in undamaged DNA, the canonical nucleotides adopt the anti orientation about the glycosidic bond (χ = ∠(O4′C1′NXCY) = 180 ± 90°, with NXCY = N1C2 for pyrimidines and N9C4 for purines), which presents the WC hydrogen-bonding face to the opposing strand (Figure 1). In contrast, damage to the nucleobase can lead to more favorable interactions with the H binding face of

Figure 2. B3LYP/6-31G(d,p) optimized A:T and G:C canonical WC base pairs. Select geometrical parameters (distances in Å and angles in deg, in parentheses) and B3LYP-D3(BJ)/6-311++G(2df,2p) binding energies (kJ mol−1, in square brackets) are provided.

the undamaged A:T pair contains two strong hydrogen bonds, while the G:C pair is held together by three strong interactions. In both complexes, the interplanar angle is near 0°, the interstrand distance (R(C1′···C1′)) is ∼10.6−10.8 Å, and the base-pair opening as measured from the purine (∠(N9C1′C1′)) is ∼55°. The interaction energy for the G:C pair (−117.2 kJ mol−1) is greater than that for A:T (−62.6 kJ mol−1), with both values being in good agreement with benchmark data previously obtained using CCSD(T) at the complete basis set limit (−128.9 and −70.3 kJ mol−1, respectively).47 When all combinations of hydrogen-bonding interactions between the WC or H face of each of the seven N-methylated nucleobases and the WC or H face of each of the four canonical nucleobases were considered (Figure 1), 105 stable base pairs were characterized. Although the structures and binding C

DOI: 10.1021/acs.jpca.7b10485 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A energies for all hydrogen-bonded complexes are provided in the Supporting Information, only the best pair for each nucleobase combination based on comparison to the structural and energetic data for the canonical pairs (Figure 2) is presented in the main text. In each case, the influence of the methyl substituent on the corresponding canonical A:T or G:C WC base pair is considered, as well as the structure and stability of the associated base mispairs. As possible, our structures are assessed against data available from experimental NMR or Xray crystal structures.19,21−23 The implications of our findings to the previously reported AlkB repair efficiency12,14 and mutational profile11,20 are discussed for each lesion where applicable. In the sections below, we initially focus on lesions in which the methyl group was added to the exocyclic amino group (2MeG, 6MeA, and 4MeC). Subsequently, neutral lesions formed upon addition of the methyl group to the nucleobase ring are examined (1MeG and 3MeT), followed by the related cationic lesions (1MeA and 3MeC). Nucleobase Lesions formed through Methylation of an Exocyclic Amino Group. N2-Methylguanine. In total, 20 base pairs were characterized between 2MeG and the canonical nucleobases (Figures 3, S1, and S2). When the additional methyl group is directed away from the WC face of G (Figures 3a and S1), 2MeG maintains three strong hydrogen bonds with C, with deviations in the hydrogen-bond distances (angles) from the canonical G:C pair being less than 0.009 Å (4°; Figure 2). Furthermore, 2MeG:C exhibits an interplanar angle (0.0°), base-pair width (10.772 Å), and base-pair opening (53.2°) that are consistent with undamaged DNA pairs. As a result, the formation of 2MeG does not affect the stability of the G:C pair (interaction energy increases by 0.9 kJ mol−1). Although stable base pairs containing two hydrogen bonds are also formed between the WC face of 2MeG and the WC face of T or the H face of A, these combinations are ∼50 kJ mol−1 less stable than 2MeG:C. In addition, 2MeG:T exhibits distortion in the shear base-pair parameter, while 2MeG:A is slightly buckled (interplanar angle = 27.7°). Furthermore, although the H face of 2MeG can similarly form two strong hydrogen bonds with the WC face of G, this pair is highly twisted (interplanar angle = 52.5°). Thus, the additional methyl substituent can be accommodated in a way that permits 2MeG to maintain a basepairing preference for a complementary C. Although the (2MeG)O6···H−N4(C) hydrogen bond is maintained when the additional methyl group is directed toward the WC face of the lesion (Figure 3b), the (2MeG)N1−H···N3(C) interaction distance increases by ∼0.4 Å, and the (2MeG)N2−H···O2(C) contact is fully disrupted, leading to an interplanar angle of 30.7°. Nevertheless, WC-2MeG:WC-C remains the strongest and structurally most favorable pair for this lesion conformation (Figures 3b and S2). Thus, regardless of the methylamino group orientation, 2MeG exhibits the strongest base-pairing interactions with C, which directly correlates with the reported nonmutagenic nature of this methylated lesion.11 N6-Methyladenine. Among 20 unique hydrogen-bonded pairs involving two different N6-methyl conformations of 6MeA (Figures 4, S3, and S4), the most stable dimer is formed when the methyl group is directed toward the WC face of the lesion, and the H face of 6MeA interacts with the WC face of G (−70.2 kJ mol−1, Figure 4b). Although this pair is 7.6 kJ mol−1 more stable than canonical A:T and contains a base-pair width consistent with natural DNA (10.971 Å), distortion in the interplanar angle (24.1°) and base-pair opening angle (38.8°) may prevent this mispair from forming in the DNA helical

Figure 3. Preferred B3LYP/6-31G(d,p) optimized base pairs between 2MeG and the canonical nucleobases with the additional methyl group directed (a) away from or (b) toward the WC face of the lesion. Select geometrical parameters (distances in Å and angles in deg, in parentheses) and B3LYP-D3(BJ)/6-311++G(2df,2p) binding energies (kJ mol−1, in square brackets) are provided. The complete set of 2MeG base pairs can be found in Figures S1 and S2.

environment. Even greater distortion is seen in the 6MeA:G combination when the additional methyl group is directed away from the lesion WC face (Figure 4a), particularly in the propeller base-pair parameter. Excluding the 6MeA:G mispair, the WC-6MeA:WC-T pair with the N6-methyl substituent D

DOI: 10.1021/acs.jpca.7b10485 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A Figure 4. continued

(kJ mol−1, in square brackets) are provided. The complete set of 6MeA base pairs can be found in Figures S3 and S4.

directed away from the WC face of the lesion is the most stable dimer (−63.1 kJ mol−1, Figure 4a). Furthermore, the structure of the WC-6MeA:WC-T pair is very similar to that of A:T (Figure 2), with deviations in the hydrogen-bond distances and angles less than 0.024 Å and 2.7°, respectively, which matches previous computational models of the 6MeA:T WC base pair.29 When the methyl group is directed toward the WC face of 6MeA (Figures 4b and S4), the pairing with T is disrupted, leading to a decreased R(C1′···C1′) distance (9.918 Å) and a nonplanar dimer (interplanar angle = 34.3°). All other base pairs are less stable than WC-6MeA:WC-T by ∼25−30 kJ mol−1. Nevertheless, the structure of the most stable WC6MeA:WC-T matched dimer identified in the present work correlates with a crystal structure of DNA containing 6MeA:T based on a cross-linked protein−DNA system21 and the experimentally reported lack of mutagenic behavior for 6MeA.11 This highlights the value of investigating the basepairing properties of methylated lesions with DFT and nucleobase dimer models. N4-Methylcytosine. Because of the smaller ring size of the pyrimidines compared to the purines, there are fewer pairing combinations available for methylated pyrimidines that maintain the basic structural features of canonical base pairs in DNA duplexes. Specifically, among the 16 characterized 4MeC hydrogen-bonded dimers (Figures S5−S6), no 4MeC:pyrimidine pairs are suitable within the DNA duplex context. Indeed, the interaction energies of 4MeC:pyrimidine pairs are as low as ca. −15 kJ mol−1, and the base-pair widths range from ∼8.2 to 12.6 Å (Figures 5, S5, and S6). Although a stable WC4MeC:WC-A dimer with the methyl group directed away from the WC face of the lesion can be identified that contains a strong (4MeC)N4−H···N1(A) interaction and has a total interaction energy of −30.7 kJ mol−1, this mispair is nonplanar (interplanar angle = 19.4°), and exhibits displacement in the base-pair shear and opening parameters (Figure 5a). No stable 4MeC:A pair could be identified when the methyl substituent is directed toward the WC face of the lesion (Figures 5b and S6). In contrast to the lesion mispairs, although N4-methylation directed toward the WC face of C lengthens the (4MeC)N3··· H−N1(G) contact compared to canonical G:C (by 0.257 Å), and repulsion between the N4-methyl substituent and O6 of G leads to propeller distortion (interplanar angle = 47.6°), the strong (4MeC)O2···H−N2(G) interaction of G:C is maintained (distance within 0.03 Å and angle within 10°), which leads to a highly stable WC-4MeC:WC-G pair (−81.0 kJ mol−1). Nevertheless, the most stable 4MeC pairing combination involves the lesion orientation that directs the N4-methyl substituent away from the WC face of the opposing G (Figure 5a), which maintains the structural features of the canonical G:C WC pair (all hydrogen-bonding interactions are within 0.013 Å and 5.2°) and a large interaction energy (−120.4 kJ mol−1). This structure parallels that reported for the 4MeC:G base pairs in a crystal structure of (5′-CGCG(4MeC)G)2,23 further highlighting the usefulness of our DFT approach. Regardless, as discussed for 2MeG, methylation at N4 of C does not affect the preference of canonical C for G. Therefore, the significantly greater stability and favorable structural parameters for 4MeC:G compared to 4MeC mispairs correlates

Figure 4. Preferred B3LYP/6-31G(d,p) optimized base pairs between 6MeA and the canonical nucleobases with the additional methyl group directed (a) away from or (b) toward the WC face of the lesion. Select geometrical parameters (distances in Å and angles in deg, in parentheses) and B3LYP-D3(BJ)/6-311++G(2df,2p) binding energies E

DOI: 10.1021/acs.jpca.7b10485 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

67.7 kJ mol −1 , Figures 2, 6, and S7). Although the (1MeG)O6···H−N4(C) interaction is maintained upon N1-

Figure 6. Preferred B3LYP/6-31G(d,p) optimized base pairs between 1MeG and the canonical nucleobases. Select geometrical parameters (distances in Å and angles in deg, in parentheses) and B3LYP-D3(BJ)/ 6-311++G(2df,2p) binding energies (kJ mol−1, in square brackets) are provided. The complete set of 1MeG base pairs can be found in Figure S7.

methylation, the other two strong hydrogen bonds in canonical G:C are disrupted. As a result, the base-pair opening changes significantly such that the 1MeG:C pair is more than 2 Å wider than conventional DNA base pairs, which would lead to significant helical disruption if this structure persists in DNA duplexes. Although a second 1MeG:C WC binding arrangement (WC-1MeG:WC-C-2, Figure S7) maintains (1MeG)O6···H−N4(C) and (1MeG)N2−H···O2(C) interactions by adopting an interplanar angle of 79.8° to alleviate steric repulsion between the N1-methyl substituent of G and N3 of C, the resulting base pair width is too short (9.150 Å), and the buckle renders this pairing geometry unfeasible in a DNA duplex. Thus, methylation at N1 of G significantly disrupts the canonical G:C pair, which likely contributes to the observed mutagenicity of this lesion.20 The structural distortion of the 1MeG:C base pair may also contribute to the repair efficiency that AlkB exhibits toward 1MeG when paired with C.14 Nevertheless, it is possible that a combination of the structure imposed by the sugar−phosphate backbone and stacking interactions with flanking bases could stabilize the isolated 1MeG:C binding arrangements characterized in the present work and thereby rationalize the albeit minimal successful 1MeG replication reported in the literature.20 Among the eight characterized hydrogen-bonded 1MeG mispairs (Figure S7), the most structurally and energetically stable 1MeG mispairs involve the H face of the lesion (Figure 6). The strongest 1MeG hydrogen-bonding interaction occurs with the WC face of G (−73.0 kJ mol−1). However, both 1MeG:G dimers characterized are severely buckled (Figure S7),

Figure 5. Preferred B3LYP/6-31G(d,p) optimized base pairs between 4MeC and the canonical nucleobases with the additional methyl group directed (a) away from or (b) toward the WC face of the lesion. Select geometrical parameters (distances in Å and angles in deg, in parentheses) and B3LYP-D3(BJ)/6-311++G(2df,2p) binding energies (kJ mol−1, in square brackets) are provided. The complete set of 4MeC base pairs can be found in Figures S5 and S6.

with the experimentally observed nonmutagenic replication of 4MeC.11 Neutral Nucleobase Lesions Formed through Methylation at a Ring Nitrogen on the WC Hydrogen-Bond Face. N1-Methylguanine. Unlike 2MeG, the formation of 1MeG directly affects the WC binding face of canonical G. As a result, the WC-1MeG:WC-C pair is less stable than G:C (by F

DOI: 10.1021/acs.jpca.7b10485 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A with interplanar angles of 62.3 or 79.4°. The interaction energies for the 1MeG pairs with A and T are similar (ranging between ca. −30 and −40 kJ mol−1 regardless of the binding arrangement, Figure S7). In both cases, select geometrical features of the 1MeG mispair may not be favorable for the DNA environment. Specifically, the base-pair opening is too narrow for 1MeG:A (22.0°), while the base-pair width is too small for 1MeG:T (7.656 Å, Figure 6). If the corresponding angle or distance increases in DNA duplexes, the calculated interaction energy will be further weakened, underscoring that these mispairs are not ideal for the helical environment. These structural data correlate with the appreciable repair of 1MeG mispairs by AlkB previously reported in the literature.14 Regardless, our small model calculations predict A and T to be the preferential pairing partners for 1MeG. This finding is in agreement with experimental data that suggests 1MeG can result in G → T transversion and G → A transition mutations,20 although our small model studies cannot explain the observed greater frequency of the transversion mutations. N3-Methylthymine. As discussed for 1MeG, addition of a methyl group to N3 of T disrupts the WC face, and results in only one strong (3MeT)O4···H−N6(A) interaction with A (hydrogen-bond distance >1.943 Å and angle 75°, Figures 9 and S10), with the exception of the WC-3MeC:H-A combination, which has a large base-pair width (R(C1′···C1′) = 11.712 Å) and base-pair opening (∠(N1C1′C1′) = 16.6°). As discussed for 3MeT, these structural features highlight the significant impact of pyrimidine methylation at a ring nitrogen on the WC face on canonical DNA base pairing. Regardless, the calculated distorted dimers irrespective of the opposing base support the observation that 3MeC is sometimes paired with G upon replication, but it also leads to C → T transition and C → A transversion mutations upon replication.20 Furthermore, the calculated distorted pairs correlates with 3MeC being one of the best AlkB targets, regardless of whether a complementary or mispairing base is considered.12,14 Future large-scale modeling initiated from the structures reported herein will be vital to further probe the weakly mutagenic behavior of 3MeC.

■

CONCLUSIONS In the present study, density functional theory was used to investigate potential hydrogen-bonded dimers between Nmethylated and canonical DNA nucleobases. The validity of our predicted structures is confirmed based on similarities to those characterized by experimental X-ray crystal and NMR studies for select lesion-pairing combinations. By evaluating the structure and stability of all dimers between the complete set of N-methylated nucleobases with modified WC binding faces and each canonical DNA base, our data reveal the effects of H

DOI: 10.1021/acs.jpca.7b10485 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

cytosine; 6MeA, N6-methyladenine; A, adenine; BJ, BeckeJohnson; C, cytosine; DFT, density functional theory; ds, double-stranded; G, guanine; H, Hoogsteen; ss, singlestranded; T, thymine; WC, Watson−Crick

nucleobase methylation on the base-pairing potential are highly dependent on the damage site and nucleobase considered. Specifically, methylation at an exocyclic amino group does not disrupt canonical WC hydrogen bonding, which correlates with the experimental observation that such lesions are nonmutagenic. Indeed, even when the methyl substituent at the exocyclic amino group is directed toward the WC binding face of the lesion, a clear preference for canonical complementary base pairing prevails. In contrast, N-methylation directly on the nucleobase ring at the WC binding face significantly disrupts canonical WC pairing and leads to distorted and destabilized base mispairs. In most cases, the structures and relative stabilities of the isolated nucleobase dimers rationalize the mutational outcomes and correlate with observed AlkB repair efficiencies. However, reasons for the observed biological outcomes for other lesions are not clear. In these cases, it is highly likely that environmental effects within the DNA duplex and/or enzyme active sites play significant roles in the observed outcomes. The fundamental information gathered from the present work about the possible hydrogen-bonding patterns between N-methylated lesions and canonical bases is critical for further probing the detrimental effects of DNA methylation using large-scale modeling of damaged DNA helices and their interactions with polymerase and repair enzymes.

■

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpca.7b10485. All characterized B3LYP/6-31G(d,p) base pairs involving the canonical nucleobases and 2MeG, 6MeA, 4MeC, 1MeG, 3MeT, 1MeA, and 3MeC. Full citations for references 13 and 45 (PDF)

■

REFERENCES

(1) Cadet, J.; Davies, K. J. A.; Medeiros, M. H. G.; Di Mascio, P.; Wagner, J. R. Formation and repair of oxidatively generated damage in cellular DNA. Free Radical Biol. Med. 2017, 107, 13−34. (2) Burney, S.; Caulfield, J. L.; Niles, J. C.; Wishnok, J. S.; Tannenbaum, S. R. The chemistry of DNA damage from nitric oxide and peroxynitrite. Mutat. Res., Fundam. Mol. Mech. Mutagen. 1999, 424, 37−49. (3) Soll, J. M.; Sobol, R. W.; Mosammaparast, N. Regulation of DNA akylation damage repair: Lessons and therapeutic opportunities. Trends Biochem. Sci. 2017, 42, 206−218. (4) Sedgwick, B. Repairing DNA-methylation damage. Nat. Rev. Mol. Cell Biol. 2004, 5, 148−157. (5) Kataoka, H.; Yamamoto, Y.; Sekiguchi, M. A new gene (AlkB) of Escherichia-coli that controls sensitivity to methyl methane sulfonate. J. Bacteriol. 1983, 153, 1301−1307. (6) Margison, G. P.; Margison, J. M.; Montesano, R. Methylated purines in deoxyribonucleic-acid of various syrian-golden-hamster tissues after administration of a hepatocarcinogenic dose of dimethylnitrosamine. Biochem. J. 1976, 157, 627−634. (7) Sedgwick, B.; Bates, P. A.; Paik, J.; Jacobs, S. C.; Lindahl, T. Repair of alkylated DNA: Recent advances. DNA Repair 2007, 6, 429− 442. (8) Luo, G.-Z.; Blanco, M. A.; Greer, E. L.; He, C.; Shi, Y. DNA N6methyladenine: A new epigenetic mark in eukaryotes? Nat. Rev. Mol. Cell Biol. 2015, 16, 705−710. (9) Yu, M.; Ji, L.; Neumann, D. A.; Chung, D. H.; Groom, J.; Westpheling, J.; He, C.; Schmitz, R. J. Base-resolution detection of N4methylcytosine in genomic DNA using 4mC-Tet-assisted-bisulfitesequencing. Nucleic Acids Res. 2015, 43, No. e148. (10) Jeggo, P. A.; Pearl, L. H.; Carr, A. M. DNA repair, genome stability and cancer: A historical perspective. Nat. Rev. Cancer 2015, 16, 35−42. (11) Fedeles, B. I.; Singh, V.; Delaney, J. C.; Li, D.; Essigmann, J. M. The AlkB family of Fe(II)/α-ketoglutarate dependent dioxygenases: Repairing nucleic acid alkylation damage and beyond. J. Biol. Chem. 2015, 290, 20734−20742. (12) Yu, B.; Hunt, J. F. Enzymological and structural studies of the mechanism of promiscuous substrate recognition by the oxidative DNA repair enzyme AlkB. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 14315−14320. (13) Chang, S. C.; Fedeles, B. I.; Wu, J.; Delaney, J. C.; Li, D.; Zhao, L.; Christov, P. P.; Yau, E.; Singh, V.; Jost, M.; et al. Next-generation sequencing reveals the biological significance of the N2,3-ethenoguanine lesion in vivo. Nucleic Acids Res. 2015, 43, 5489−5500. (14) Chen, F.; Tang, Q.; Bian, K.; Humulock, Z. T.; Yang, X.; Jost, M.; Drennan, C. L.; Essigmann, J. M.; Li, D. Adaptive response enzyme AlkB preferentially repairs 1-methylguanine and 3-methylthymine adducts in double-stranded DNA. Chem. Res. Toxicol. 2016, 29, 687−693. (15) Li, D. Y.; Delaney, J. C.; Page, C. M.; Yang, X. D.; Chen, A. S.; Wong, C. Y.; Drennan, C. L.; Essigmann, J. M. Exocyclic carbons adjacent to the N6 of adenine are targets for oxidation by the Escherichia coli adaptive response protein AlkB. J. Am. Chem. Soc. 2012, 134, 8896−8901. (16) Li, D. Y.; Fedeles, B. I.; Shrivastav, N.; Delaney, J. C.; Yang, X. D.; Wong, C. Y.; Drennan, C. L.; Essigmann, J. M. Removal of N-alkyl modifications from N2-alkylguanine and N4-alkylcytosine in DNA by the adaptive response protein AlkB. Chem. Res. Toxicol. 2013, 26, 1182−1187. (17) Shrivastav, N.; Li, D. Y.; Essigmann, J. M. Chemical biology of mutagenesis and DNA repair: Cellular responses to DNA alkylation. Carcinogenesis 2010, 31, 59−70.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: (403) 329-2323. Fax: (403) 329-2057. ORCID

Stefan A. P. Lenz: 0000-0003-4951-9578 Stacey D. Wetmore: 0000-0002-5801-3942 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank the Natural Sciences and Engineering Research Council of Canada (NSERC, 2016-04568), Canada Research Chairs Program (950-228175), and Canada Foundation for Innovation (22770) for financial support. S.A.P.L. acknowledges NSERC (CGS-D), Alberta Innovates−Technology Futures (AI-TF), and the University of Lethbridge, while L.R.F. acknowledges NSERC (USRA) for student scholarships. Computational resources from the New Upscale Cluster for Lethbridge to Enable Innovative Chemistry (NUCLEIC) and those provided by Westgrid and Compute/Calcul Canada are greatly appreciated.

■

ABBREVIATIONS 1MeA, N1-methyladenine; 1MeG, N1-methylguanine; 2MeG, N2-methylguanine; 3MeC, N3-methylcytosine; 3MeT, N3methylthymine; 4MeC, N4-methylcytosine; 5MeC, N5-methylI

DOI: 10.1021/acs.jpca.7b10485 J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A (18) Yang, H.; Lam, S. L. Effect of 1-methyladenine on thermodynamic stabilities of double-helical DNA structures. FEBS Lett. 2009, 583, 1548−1553. (19) Yang, H.; Zhan, Y.; Fenn, D.; Chi, L. M.; Lam, S. L. Effect of 1methyladenine on double-helical DNA structures. FEBS Lett. 2008, 582, 1629−1633. (20) Delaney, J. C.; Essigmann, J. M. Mutagenesis, genotoxicity, and repair of 1-methyladenine, 3-alkylcytosines, 1-methylguanine, and 3methylthymine in AlkB Escherichia coli. Proc. Natl. Acad. Sci. U. S. A. 2004, 101, 14051−14056. (21) Lu, L.; Yi, C.; Jian, X.; Zheng, G.; He, C. Structure determination of DNA methylation lesions N1-meA and N3-meC in duplex DNA using a cross-linked protein-DNA system. Nucleic Acids Res. 2010, 38, 4415−4425. (22) Sathyamoorthy, B.; Shi, H.; Zhou, H.; Xue, Y.; Rangadurai, A.; Merriman, D. K.; Al-Hashimi, H. M. Insights into Watson-Crick/ Hoogsteen breathing dynamics and damage repair from the solution structure and dynamic ensemble of DNA duplexes containing m1A. Nucleic Acids Res. 2017, 45, 5586−5601. (23) Cervi, A. R.; Guy, A.; Leonard, G. A.; Teéoule, R.; Hunter, W. N. The crystal structure of N4-methylcytosine. Guanosine base-pairs in the synthetic hexanucleotide d(CGCGm4CCG). Nucleic Acids Res. 1993, 21, 5623−5629. (24) Forde, G.; Flood, A.; Salter, L.; Hill, G.; Gorb, L.; Leszczynski, J. Theoretical ab initio study of the effects of methylation on structure and stability of G:C Watson-Crick base pair. J. Biomol. Struct. Dyn. 2003, 20, 811−817. (25) Flood, A.; Hubbard, C.; Forde, G.; Hill, G.; Gorb, L.; Leszczynski, J. Theoretical ab initio study of the effects of methylation on the nature of hydrogen bonding in A:T base pair. J. Biomol. Struct. Dyn. 2003, 21, 297−302. (26) Aliakbar Tehrani, Z.; Fattahi, A. Comparison of gas phase intrinsic properties of cytosine and thymine nucleobases with their Oalkyl adducts: Different hydrogen bonding preferences for thymine versus O-alkyl thymine. J. Mol. Model. 2013, 19, 2993−3005. (27) Qiu, Z. M.; Wang, H. L.; Liu, Y. Z.; Hou, D. N. MP2 study on the hydrogen-bonding interaction between O4-methylthymine and DNA bases: A, C, G, and T. Struct. Chem. 2014, 25, 767−774. (28) Acosta-Silva, C.; Branchadell, V.; Bertran, J.; Oliva, A. Mutual relationship between stacking and hydrogen bonding in DNA. Theoretical study of guanine−cytosine, guanine−5-methylcytosine, and their dimers. J. Phys. Chem. B 2010, 114, 10217−10227. (29) Villani, G. Effect of methylation on the properties of the Hbridges in DNA. A systematic theoretical study on the couples of base pairs. J. Phys. Chem. B 2015, 119, 7931−7943. (30) Wibowo, F. R.; Trieb, M.; Rauch, C.; Wellenzohn, B.; Liedl, K. R. The N6-methyl group of adenine further increases the BI stability of DNA compared to C5-methyl groups. J. Phys. Chem. B 2005, 109, 557−564. (31) Karolak, A.; van der Vaart, A. BII stability and base step flexibility of N6-adenine methylated GATC motifs. Biophys. Chem. 2015, 203−204, 22−27. (32) Portella, G.; Battistini, F.; Orozco, M. Understanding the connection between epigenetic DNA methylation and nucleosome positioning from computer simulations. PLoS Comput. Biol. 2013, 9, No. e1003354. (33) Millen, A. L.; Churchill, C. D. M.; Manderville, R. A.; Wetmore, S. D. Effect of Watson−Crick and Hoogsteen base pairing on the conformational stability of C8-phenoxyl-2′-deoxyguanosine adducts. J. Phys. Chem. B 2010, 114, 12995−13004. (34) Hokmabady, L.; Raissi, H.; Khanmohammadi, A. Interactions of the 5-fluorouracil anticancer drug with DNA pyrimidine bases: A detailed computational approach. Struct. Chem. 2016, 27, 487−504. (35) Palafox, M. A. Anticancer drug IUdR and other 5-halogen derivatives of 2’-deoxyuridine: Conformers, hydrates, and structureactivity relationships. Struct. Chem. 2014, 25, 53−69. (36) Qiu, Z.; Xia, Y.; Wang, H.; Diao, K. Hydrogen-bonding interaction of urea with DNA bases: A density functional theory study. J. Struct. Chem. 2011, 52, 462−470.

(37) Qiu, Z. M.; Cai, H. Z.; Wang, H. L.; Xia, Y. M.; Wang, H. J. DFT and MP2 investigations on the hydrogen bonding interaction between 5,6-dihydrothymine and DNA bases: A, C, G. J. Struct. Chem. 2012, 53, 1037−1045. (38) Basumatary, J.; Bezbaruah, B.; Kalita, R.; Barman, T. K.; Medhi, C. Quantum mechanical studies on the existence of AC mismatches through prototopic tautomerization pathway in adenine and cytosine recognition. J. Theor. Comput. Chem. 2017, 16, 1750027. (39) Brovarets, O. O.; Voiteshenko, I. S.; Perez-Sanchez, H.; Hovorun, D. M. A QM/QTAIM research under the magnifying glass of the DPT tautomerisation of the wobble mispairs involving 2aminopurine. New J. Chem. 2017, 41, 7232−7243. (40) Masoodi, H. R.; Bagheri, S.; Abareghi, M. The effects of tautomerization and protonation on the adenine-cytosine mismatches: A density functional theory study. J. Biomol. Struct. Dyn. 2016, 34, 1143−1155. (41) Dai, Q.; Sanstead, P. J.; Peng, C. S.; Han, D. L.; He, C.; Tokmakoff, A. Weakened N3 hydrogen bonding by 5-formylcytosine and 5-carboxylcytosine reduces their base-pairing stability. ACS Chem. Biol. 2016, 11, 470−477. (42) Grimme, S.; Ehrlich, S.; Goerigk, L. Effect of the damping function in dispersion corrected density functional theory. J. Comput. Chem. 2011, 32, 1456−1465. (43) Merrick, J. P.; Moran, D.; Radom, L. An evaluation of harmonic vibrational frequency scale factors. J. Phys. Chem. A 2007, 111, 11683− 11700. (44) Boys, S. F.; Bernardi, F. The calculation of small molecular interactions by the differences of separate total energies. Some procedures with reduced errors. Mol. Phys. 1970, 19, 553−566. (45) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; et al. Gaussian 09, Revision D.01; Gaussian, Inc.: Wallingford, CT, 2016. (46) Šponer, J.; Šponer, J. E.; Mládek, A.; Banás,̌ P.; Jurečka, P.; Otyepka, M. How to understand quantum chemical computations on DNA and RNA systems? A practical guide for non-specialists. Methods 2013, 64, 3−11. (47) Rejnek, J.; Hobza, P. Hydrogen-bonded nucleic acid base pairs containing unusual base tautomers: Complete basis set calculations at the MP2 and CCSD(T) levels. J. Phys. Chem. B 2007, 111, 641−645.

J

DOI: 10.1021/acs.jpca.7b10485 J. Phys. Chem. A XXXX, XXX, XXX−XXX