Abstract
The reaction of hydroxyl radical (OH•) with DNA accounts for about half of radiation-induced DNA damage in living systems. Previous literature reports point out that the reaction of OH• with DNA proceeds mainly through the addition of OH• to the C=C bond of the DNA bases. However, recently it has been reported that the principal reaction of OH• with dGuo (deoxyguanosine) is the direct hydrogen atom abstraction from its exocyclic amine group rather than addition of OH• to the C=C bond. In the present work, these two reaction pathways of OH• attack on guanine (G) in the presence of water molecules (aqueous environment) are investigated using the density functional theory (DFT) B3LYP method with 6-31G* and 6-31++G** basis sets. The calculations show that the initial addition of the OH• at C4=C5 double bond of guanine is barrier free and the adduct radical (G-OH•) has only a small activation barrier of ca. 1 – 6 kcal/mol leading to the formation of a metastable ion-pair intermediate (G•+---OH−). The formation of ion-pair is a result of the highly oxidizing nature of the OH• in aqueous media. The resulting ion-pair (G•+---OH−) deprotonates to form H2O and neutral G radicals favoring G(N1-H)• with an activation barrier of ca. 5 kcal/mol. The overall process from the G(C4)-OH• (adduct) to G(N1-H)• and water is found to be exothermic in nature by more than 13 kcal/mol. (G-OH•), (G•+---OH−), and G(N1-H)• were further characterized by the CAM-B3LYP calculations of their UV-visible spectra and good agreement between theory and experiment is achieved. Our calculations for the direct hydrogen abstraction pathway from N1 and N2 sites of guanine by the OH• show that this is also a competitive route to produce G(N2-H)•, G(N1-H)• and H2O.
Keywords: OH radical, ion-pair formation, radical cation, G•+, neutral radical, guanine, oxidation, excited states of DNA radical, DFT, B3LYP, OH-radical adduct
Introduction
Exposure of living systems to ionizing radiation generates damage by the ionization of DNA (direct effect), ionization of the first hydration shell that transfers damage to DNA (quasi direct effect) and by ionization of water (indirect effect) which produces OH•, H• and aqueous electron (e−aq).1,2 Of the water radical species it is the reaction of OH• with DNA has been found to be the most damaging as it accounts for over half of radiation damage to cellular systems by low LET (linear energy transfer) radiation.1,2 For this reason the reactions of OH• with DNA components including DNA bases, nucleosides and nucleotides have been extensively studied with a variety of experimental techniques.3–8 O’Neill and co-workers8 studied the reaction of OH• with guanine and adenine in Guo and in Ado using pulse radiolysis and the reaction of OH• was reported to be addition reactions at C4, C5 and C8 atoms of adenine and guanine.3–5,8 Later, using pulse radiolysis with optical and conductance detection the reaction of OH• with guanine and adenine were extensively explored by Steenken, Vieira and Candeias5,9. They reported that addition of OH• at C4 or C5 in adenine or guanine leads to neutral radical formation ((A-H)• and G(N1-H)•) through dehydration reaction while addition at C8 results in the formation of a highly reducing radical, i.e., 8-hydroxy-7,8-dihydroguanin-7-yl-radical, that on further oxidation leads to 8-oxo-G or on reduction to Fapy-G.3,4 They, also, proposed that in monomers, the addition reaction of OH• with guanine at its C4=C5 dominates (~60 – 70%) over addition to its C8 (~17%).9g We note here that this relative small extent of addition of OH• to the C8 of guanine in monomers has been further supported by the small yield (~ 10%) of OH• induced formation of 8-oxo-G and Fapy-G in gamma-irradiated N2O-saturated solution of dGuo.4b They further proposed that the addition of OH• to either C4 or C5 atoms of guanine (G-OH•) lead to the radical cation and hydroxyl anion (G•+ + OH−) and/or the formation of the neutral deprotonated guanine radical (G(N1-H)•) and water (see scheme 3 in ref. 9g), as given in eq. (1a, b)
(1a) |
(1b) |
Using theory, addition reactions of OH• with guanine and adenine were reported by Colson and Sevilla,10a–c Eriksson and coworkers,10d,e Mishra and coworkers,10f–h Naumov and Sonntag10i and Mundy and coworkers.13 Very recently, the hydrogen abstraction reaction by OH• from sugar moiety in hydrated DNA was studied using molecular dynamics simulation.14
Recently, using pulse radiolysis with optical detection, the reaction of OH• with guanine has been reevaluated by Chatgilialoglu and coworkers.11 It was proposed that the main reaction of OH• with guanine is the formation of G(N2-H)• via direct hydrogen abstraction from the exocyclic NH2 group of guanine moiety by OH• which after tautomerization forms G(N1-H)• as shown in eq. 2a, b
(2a) |
(2b) |
In keeping with Chatgilialoglu’s interpretation that H-atom abstraction processes dominate and using the B3LYP/DZP++ level of theory, Cheng et al.12 recently studied the OH•-induced hydrogen atom abstraction reaction from the different sites of adenine. Direct hydrogen atom abstraction reaction from guanine’s exocyclic NH2 group by OH• in the gas-phase using Car–Parinello molecular dynamics (MD) was also reported by Mundy et al.13a However, in a later study from this group, the dehydrogenation reaction from four sites (N1, N2, N9 and C8) of guanine by OH• was carried out in an aqueous phase of many explicit water molecules.13b For the reaction of OH• at NH2 of guanine, they suggested a possible transient ion-pair (G•+---OH−) intermediate that subsequently underwent deprotonation from the NH2 group of transiently formed guanine radical cation within a few picoseconds.13b It is interesting to note that Steenken also suggested a polar ion-pair type transition state for the dehydration of the C4-OH adduct of adenine.5 In this work, we perform DFT calculations which also predict such a metastable ion-pair formation after OH• addition to the C4=C5 double bond in the guanine ring (eq. (1a)).
From the review of the literature described above, the mechanism of action of OH radical with DNA, which has been a focus of study for 30 years, has become somewhat controversial. The fundamental importance of the reactions of aqueous OH radical with DNA in living systems suggests that this issue needs careful consideration by theory. The earlier studies employing theory treated systems in the gas-phase or in aqueous medium through the use of the solvation models that did not include discrete water molecules and none considered the full reaction pathway of addition/elimination vs. abstraction. In the present study, the reaction of OH• with guanine is modeled in the presence of explicit water molecules and both addition/elimination and hydrogen abstraction reactions by OH• as given in eqs. (1) and (2) are considered. Our results show that both pathways, addition/elimination and abstraction reactions of OH•, are likely with the relative yields from each pathway dependent on kinetic considerations.
Method of Calculations
The interaction of OH radical with guanine at different sites has been considered in the gas phase and in the full solvation through the use of IEF-PCM model (integral equation formalism-polarized continuum model) using the B3LYP/6-31G* method. Our calculations show that the reaction of OH• with guanine depends on the mode of approach of OH• with guanine. Interaction of OH• above the plane of the guanine ring at C4=C5 bond shows the possibility of ion-pair formation in addition to C4 and C5 adducts formation, see the results and discussion section. The relative stabilities of the different complexes formed due to OH radical interaction with guanine are shown in Figure 1. This preliminary result sets the stage to explore the reaction of OH• with guanine in the presence of explicit water molecules.
Figure 1.
Relative stabilities of C4, C5, and C8 adducts and N1-H complex formed on the interaction of guanine and OH radical in gas phase (column 1), PCM model (column 2) and in the presence of 12 explicit water molecules and in PCM (single point calculation) (column3). B3LYP/6-31G* method was used for the calculation. Pink circle shows the location of OH in the hydrated guanine.
The addition reaction of OH• at C4=C5 double bond with guanine solvated by twelve explicit water molecules is explored using the B3LYP method and 6-31G* and 6-31++G** basis sets. The structures of the adduct formation (G-OH• + 12H2O) (reactant), ion-pair (G•+---OH− + 12H2O) (intermediate) and the radicals (G(N1-H)• + 13H2O and (G(N2-H)• + 13H2O)) (product) were fully optimized using the B3LYP method and 6-31G* and 6-31++G** basis sets. The reaction profile of the formation of neutral radical (G(N1-H)• or G(N2-H)•) from the adduct (G-OH• + 12H2O) proceeds on two surfaces which are investigated as follows: (i) The potential energy surface (PES) of the formation of ion-pair from the adduct (G-OH• + 12H2O) is calculated by stretching the C4-OH and C5-OH bonds from its equilibrium bond lengths up to 2.5 Å. and (ii) the PES for the radical formation (G(N1-H)• or G(N2-H)•) is calculated by stretching the N1-H and N2-H bonds of guanine from the ion-pair. In steps (i) and (ii) the bond dissociation was carried out in the step size of 0.05 Å and at each chosen point on the PES the structures were fully optimized by constraining only the corresponding bond length, see Figure 2. The maxima or the highest points on the PES (Figure 2), which connect the reactant and product or intermediate, are estimates of the barrier heights; however, frequency calculations were not performed to verify a true transition states as this is not practical for such large systems. Relative energy (ΔE in kcal/mol) involved in the reaction along with the optimized structures of reactant (adduct), intermediate (ion-pair) and product (radical) is shown in Figure 2. The direct hydrogen abstraction reaction (eq. (2)) was also considered by placing the OH radical in the plane of the guanine ring near the N1-H and N2-H sites in the presence of twelve water molecules and PES for hydrogen abstraction from both the sites (N1 and N2) was calculated, as described above, using the B3LYP/6-31G* method. In the text, N1 refers to the N1 atom of guanine. N2 and NH2 refer to the exocyclic amine nitrogen and amine group of guanine
Figure 2.
B3LYP/6-31G* calculated potential energy surfaces (PES) for the reaction of OH radical with hydrated guanine (G + 12H2O). Reaction proceeds on each surface are depicted as C4, C5, N1 and N2. G-OH• = C4, C5 adduct; G•+---OH− = ion-pair; G(-H)• = G(N1-H)• or G(N2-H)•. For clarity only the structures of G(C4-OH)• adduct and G(N1-H)• are shown. The optimized structures of G(C5-OH)• adduct and G(N2-H)• are given in the supporting information.
The B3LYP method is widely used in several studies and found to be quite accurate in producing ground state molecular properties of DNA bases and base pairs in their neutral and radical ion states.10e–j,12,13 Considering the optimized hydrated structures of adduct, ion-pair and radical, single point energy calculations were done to take into account the effect of full solvation through the use of IEF-PCM model, see Table 1. Our calculations show that the B3LYP/6-31G* and B3LYP/6-31++G** give similar results, see Table 1, thus in the text, we discuss our results derived from B3LYP/6-31G* calculations unless otherwise stated.
Table 1.
Relative reaction energies (ΔE) in kcal/mol for OH radical addition/elimination processes with guanine solvated with 12 water molecules. Calculations were done using the B3LYP method and 6-31G* and 6-31++G** basis sets.
Speciesa | Dist. (Å) | ΔE (kcal/mol) | ||||
---|---|---|---|---|---|---|
B3LYP | ||||||
6-31G* | 6-31++G** | |||||
6-31G* | 6-31++G** | Gas phase | PCM | Gas Phase | PCM | |
G + 12 H2O + OH• | 21.8 | 20.17 | 24.38f | 21.11f | ||
G(C5)-OH• | 1.52 | 1.48b | 0.0 | 0.0c | 0.0b | 0.0e |
Barrier Heightd | 1.75 | 1.75b | 1.12 | - | 2.38b | - |
G•+--- OH− | 2.32 | 2.39b | −2.43 | −2.94c | −2.11b | −2.70e |
Barrier Heightd | 1.35 | 1.30b | 1.85 | - | 2.44b | - |
G(C4)-OH• | 1.48 | - | −2.44 | −3.55c | −4.04f | −4.90f |
G(N1-H)• | - | - | −16.40 | −18.43c | −19.38f (−18.09)b |
−21.42f (−19.90)e |
G(N2-H)• | - | - | −16.51 | −16.42c | −20.22f | −20.13f |
See Figures S2 and S3.
Values obtained from B3LYP/6-31++G** optimization.
Values obtained from single point calculations using PCM model.
Highest point on the PES.
Geometries optimized at B3LYP/6-31++G**. Values obtained from single point calculations using PCM model.
Single point calculations using B3LYP/6-31G* optimized geometries.
The presence of adduct, ion-pair and radical during the reaction were also characterized by their calculated vertical excited state UV-visible spectra using the time-dependent (TD) CAM-B3LYP method incorporating the effect of full solvation through the IEF-PCM model using the B3LYP optimized geometries. We used TD-CAMB3LYP-PCM method for excited state calculations because the B3LYP method is not appropriate for ion-pair or species involving charge transfer excited states.15 The transition energies for charge transfer states are greatly underestimated by the B3LYP method especially as the distance increases.15,16 The CAM-B3LYP method is a long-range corrected hybrid functional and found to be very successful to study the excited states of systems involving charge transfer states.16
These calculations were done using Gaussian 09 suite of programs.17 GaussView18a molecular modeling software was used to plot the molecular orbitals and draw the molecular structure. The Gabedit molecular modeling package was used to plot UV-visible spectra.18b
Results and Discussion
Guanine and OH• Interaction and Ion-Pair Formation
Adduct formation
To get information about the relative stabilities of binding of OH radical with guanine at different sites, we added OH radical at C4, C5 and C8 atoms of uanine and placed OH radical near the N1-H site in the plane of the guanine ring. The geometries thus generated were fully optimized in the gas phase as well as in the solution phase using the IEF-PCM model at the B3LYP/6-31G* level of theory. From the calculations, we found that C8 adduct is the most stable and it is stabilized by more than 34 kcal/mol relative to the reactant (G + OH•) in the gas phase and in solution (PCM). The gas phase and solution phase (PCM) calculations gave almost similar stabilization energies and has a maximum difference of ca. 2.5 kcal/mol, see Figure 1. The relative stabilities of C4, C5 (adducts) and N1-H---OH• (complex) in solution using PCM are −17.9, −11.9 and −10.28 kcal/mol, respectively, with respect to the separate G + OH• (reactant), see Figure 1. The addition of OH• at C2, C4, C5 and C8 atoms of guanine and 8-oxoguanine has been studied by Jena and Mishra10f,g using the B3LYP method and found the relative stability of the OH• adducts formation with guanine in the order C8 > C2 > C4> C5.
Ion-pair formation
Apart from the adduct formation, we also found a new structure which is an ion-pair complex, (G•+--OH−), formed after OH radical attack on guanine. In the gas phase or in solution using the PCM model only, the ion-pair is found to be more stable than the separate G + OH• but less than the OH adducts, see Figure 1. In this ion-pair complex, the OH is situated above the C4=C5 bond at a distance of 2.74 Å and 2.37 Å from C4 and C5 atoms of guanine in solution (PCM) and 2.70 Å and 2.38 Å in gas phase, respectively. To make sure that it is not the transition state structure to form either C4 or C5 adduct, we calculated the vibrational frequencies of this complex using the B3LYP/6-31G* method in the gas phase and in solution (PCM) and found all the frequencies to be positive, which confirms the formation of metastable complex between guanine and OH radical. To understand the nature of this complex, we plotted the spin density distribution of this complex, shown in Figure S1 in the supporting information. From Figure S1, we found that the unpaired spin is largely localized on the guanine, which characterizes this complex as G•+--OH− (ion-pair). From stability point of view, this ion-pair lies ca. 12 kcal/mol above the C4 adduct and 6 kcal/mol above the C5 adduct, see Figure 1.
The higher energy for the complex (ion-pair; G•+---OH−) compared to the adducts raised an important question, as to whether this ion-pair becomes more stable in the presence of explicit water molecules. To answer this question, we optimized the complex structure in the presence of 12 water molecules at the B3LYP/6-31G* level of theory and calculated the relative stabilities of the formation of ion-pair, C4 and C5 adducts and in plane N1---OH complex. The B3LYP/6-31G* calculation shows that ion-pair is stabilized by more than 2 kcal/mol with respect to the C5 adduct and it is isoenergetic to C4 adduct, see Figure 1. In the presence of 16 water molecules, ion-pair is found to be more stable than C4 or C5 adducts by 1.4 kcal/mol and 5 kcal/mol, respectively, see Figure S2 in the supporting information. Thus, these calculations suggest that in an aqueous environment the formation of ion-pair is a likely pathway. To save computational time we used the 12 water complex in our study to investigate the energetics of deprotonation reaction from N1 and N2 sites of guanine, which is discussed in the next section.
Ion-Pair Formation on OH• reaction with hydrated guanine
The addition reaction of OH radical with hydrated guanine (G + 12H2O) and subsequent processes were calculated using the B3LYP method with the 6-31G* and 6-31++G** basis sets. These results are summarized in Table 1 and discussed below.
Figure 2 shows the potential energy surface (PES) for reaction of OH radical with hydrated guanine (G + 12H2O) using the B3LYP/6-31G* method. The figure - shows the relative energies of initial reactants, i.e., G + 12H2O and OH radical, the C4 and C5 adducts (G(C4)-OH•; G(C5)-OH•), the ion-pair (G•+---OH−) and finally the radicals formed by hydrogen abstraction from N1 and N2 sites, i.e., G(N1-H)• and G(N2-H)•, respectively. The optimized geometries of G-OH• + 12H2O, G•+---OH− + 12H2O, and G(-H)• + 13H2O are shown in Figure S3 in the supporting information.
The addition reaction of OH radical at C4 or C5 atom of hydrated guanine proceeds without a barrier to G(C4)-OH• or G(C5)-OH• adduct formation and the process is exothermic by ca. 22 to 24 kcal/mol, see Figures 1 and 2. In the adduct, OH radical bonds either to C4 or C5 atom of guanine with bond lengths 1.48 Å and 1.52 Å, respectively, see Figure S3 in the supporting information. Addition of OH• at C4 is slightly energetically favored by −2.44 kcal/mol over C5 and the subsequent reaction mechanism proceeds through a transition state, see Figure 2. From the C5 adduct (G(C5)-OH• + 12H2O), the PES was scanned by detaching the OH moiety from C5 atom, shown in Figure 2. We note that after surpassing the barrier of ca. 1 kcal/mol, ion-pair formation takes place and it is more stable than the adduct (G(C5)-OH• + 12 H2O) by 2.44 kcal/mol. The B3LYP/6-31G* calculated PES for the detachment of the OH from C4 adduct (G(C4)-OH• + 12 H2O) to form ion-pair (G•+---OH− + 12H2O) has a barrier of 6.45 kcal/mol and the formation of ion-pair was found to be isoenergetic with respect to the adduct. From the ion-pair, the PES for deprotonation from N1 and N2 sites of guanine and the formation of G(N1-H)• and G(N2-H)• and water (G•+---OH− → G(-H)• + H2O (reaction 1c)) was calculated by gradually stretching the N1-H and N2-H bonds of guanine with optimization the remaining structure. The B3LYP/6-31G* calculated barrier height for deprotonation from the N1 site of guanine was calculated to be 4.28 kcal/mol with respect to the ion-pair (G•+---OH− + 12H2O). The formation of N1 radical (G(N1-H)•) was found to be highly exothermic by −16.40 kcal/mol relative to G(C5)-OH• calculated using the B3LYP/6-31G* method, see Figure 2. Using the B3LYP-PCM/6-31+G(d,p) level of theory Naumov and Sonntag10i also studied the water elimination reaction from G(C4)-OH• and G(C5)-OH• adducts to form G(N1-H)• and also found this reaction to be exothermic in nature.
The barrier for the N2-H deprotonation of guanine to form G(N2-H)• is ca. 12 kcal/mol relative to the ion-pair (G•+---OH− + 12H2O) calculated at the B3LYP/6-31G* level of theory, see Figure 2. The formation of (G(N2-H)•) is found to be exothermic by - 16.51 kcal/mol relative to G(C5)-OH• which is identical with G(N1-H)•. However, calculations employing the PCM model for full solvation of the 12 H2O system show that G(N1-H)• is found to be more stable than G(N2-H)• by 2 kcal/mol at the B3LYP-PCM/6-31G* and 4 kcal/mol calculated at the B3LYP-PCM/6-31++G** level of theory. The preferred stability of G(N1-H)• over G(N2-H)• has also been found in earlier theoretical and experimental studies.9a,19, 20
At the B3LYP/6-31++G** level of theory, the barrier height for ion-pair formation from G(C5)-OH• + 12 H2O was calculated to be 2.38 kcal/mol and the ion-pair was found to be more stable than the adduct by 2.11 kcal/mol, see Table 1. The B3LYP/6-31++G** calculated barrier height for N1 deprotonation was found to be 2.44 kcal/mol while the formation of G(N1-H)• was found to be −18 kcal/mol with respect to G(C5)-OH• + 12H2O.
The C4 or C5 adduct structure induces appreciable non-planarity in the guanine structure similar to butterfly shape as reported in earlier studies.10c,h,j As OH group leaves the guanine base to form ion-pair (G•+---OH− + 12H2O), the guanine base becomes near planar, see Figure S3(b) in the supporting information. The spin density distributions found for G-OH• + 12H2O, G•+---OH− + 12H2O and G(-H)• + 13H2O are shown in Figure S4 in the supporting information. From Figure S4, we see that the spin density distribution of the adduct (G-OH• + 12H2O) is quite different than the spin density distribution of the G•+---OH− + 12H2O (ion-pair). The plots of spin density distribution in G•+ and G•+ + 12H2O calculated in the gas-phase and in PCM model are shown in Figure S4(d) – (f). From Figure S4, it is evident that the spin density distribution in ion-pair (G•+---OH−) (Figure S4(b)) is similar to the spin density distribution in G•+ (see Figure S4(d) – (f) in the supporting information which is calculated using the B3LYP-PCM/6-31G* and B3LYP-PCM/6-31++G** methods). In addition to spin density distributions, the B3LYP/6-31++G** calculated Mulliken net charges residing on OH, G and 12H2O in (G(C5)-OH• + 12H2O) adduct in gas-phase (PCM) are: 0.014e (0.023e) for OH, −0.227e (−0.210e) for G, and 0.213e (0.187e), for 12H2O. In the ion-pair the corresponding net charges are −0.364e (−0.390e), 0.213e (0.296e) and 0.151e (0.094e) (e = electronic charge), respectively. Thus, in ion-pair, the OH moiety gains a substantial charge −0.390e and becomes anionic and guanine looses a large 0.296e charge and becomes cationic in nature. These results clearly support the formation of a metastable ion-pair (G•+---OH−) after OH addition to C4/C5 of guanine as was suggested by Mundy and coworkers for OH attack on the NH2 portion of guanine.13b Such reactions are well known for planar conjugated systems; for example, using time-resolved resonance Raman spectroscopy, Tripathi studied the addition reaction of OH• with p-dimethoxybenzene and found the cation radical formation after the loss of OH−.21
In summary, our calculations show that the reaction of OH radical with guanine has no barrier to C4 or C5 adduct formation as reported in earlier studies10f,g,i and with hydration it subsequently transformed to the more stable ion-pair. From the ion-pair (G•+---OH− + 12H2O) deprotonation reaction takes place either from N1 or N2 sites of guanine analogous to the guanine radical cation which also prefers these sites for deprotonation. The present calculation shows that the lowest barrier path for deprotonation is from N1 site of guanine (Figure 2). It is well known that G•+ becomes acidic (pKa = 3.9)9a,g and quickly deprotonates in the presence of aqueous media from N1 site as found in other studies.20,23,24 We are, also, aware that in hydrated environment a number of dynamic solvation structures are available which will affect the presently calculated PESs and the barrier heights of the reaction. Our calculations represent one of the possible local minimum structures present during solvent dynamics and can not be considered as the expected experimental barriers.
Direct hydrogen abstraction from N1H and NH2 group of guanine
The direct hydrogen abstraction reaction (eq. 2) by OH• from N1 and N2 sites of guanine to form neutral radical was also calculated using the B3LYP/6-31G* method in gas phase as well as in aqueous phase using PCM model. From our calculations we found that the in plane reaction of OH radical depends on the direction of approach of the OH radical towards guanine. Our calculations show that when OH radical approaches between N1 and NH2 region, it preferentially abstracts an hydrogen from NH2 group (in gas phase) and in the presence of aqueous solution modeled through PCM the OH radical preferentially abstracts an hydrogen atom from N1 site, see Figure S8 in the supporting information. These reactions proceed with negligible barriers to form radicals, i.e., G(N2-H)• (in gas phase) or G(N1-H)• (in PCM). If the in plane approaching OH radical is in between the O6 and N1H region of guanine, a hydrogen-bonded complex is formed as shown in Figure 1 and in Figure S8.
Another picture arises for the reactivity of the OH radical in aqueous environment with explicit water molecules. The reaction critically depends on the conformation and arrangement of the interacting water molecules and this is well documented in the literature.22 Depending on the water conformation the barrier for the hydrogen abstraction from CH4 by OH radical with various levels of solvation (0 – 3 H2O) ranged from 2.37 kcal/mol to 5.25 kcal/mol as calculated using the CCSD(T)/aug-cc-pVDZ level of theory.22 In our case, we also found that for guanine solvated by 12 waters (G + 12 H2O) the energy involved for direct hydrogen abstraction from the N1H or NH2 site by the OH radical increases from the gas phase results which had negligible barriers to 7 kcal/mol for N2 and 10 kcal/mol for N1 hydrogen abstraction, see Figure S9 in the supporting information. In this calculation, we placed OH radical near N1 and N2 sites of hydrated guanine (G + 12 H2O) and scanned the PES for hydrogen abstraction from N1 and N2 sites, see Figure S9. As stated above, this barrier likely depends on the arrangement of the specific water molecules and would change if dynamics were considered. The calculation also showed that in these arrangements the spin density is fully localized on OH until abstraction takes place (see Figure S9) which is different than the addition reaction in which the spin density is transferred to the guanine immediately on interaction, see in Figures 2 and S4.
Excited states of hydrated G-OH•, G•+---OH− and G(-H)•
In Figures 3 – 5, we show the UV-visible spectra of G-OH• + 12H2O, G•+---OH− + 12H2O, and G(N1-H)• + 13H2O calculated using the TD-CAM-B3LYP-PCM/6-31G* method incorporating the full solvent through the IEF-PCM model. The corresponding spectra computed using the TD-CAM-B3LYP-PCM/6-31++G** method are shown in the supporting information as Figures S5 – S7, respectively. In Table S1 in the supporting information, we provided transition energies along with the oscillator strengths of G•+ + 12H2O, G-OH• + 12H2O, G•+---OH− + 12H2O, G(N1-H)• + 13H2O and G(N2-H)• + 13H2O calculated using the TD-CAM-B3LYP-PCM and TD-B3LYP-PCM methods using 6-31G* and 6-31++G** basis sets. From Table S1, a good agreement between theory and experiment where available is evident.
Figure 3.
TD-CAM-B3LYP-PCM/6-31G*//B3LYP/6-31G* calculated excited states of G-OH• + 12H2O. MOs dominantly involved in the transitions are also shown.
Figure 5.
TD-CAM-B3LYP-PCM/6-31G*//B3LYP/6-31G* calculated excited states of G(N1-H)• + 13H2O. MOs dominantly involved in the transitions are also shown.
Excited state spectra of G-OH• + 12H2O (adduct), shown in Figure 3, shows three intense peaks located at 249 nm, 328 nm and 453 nm having oscillator strengths 0.1122, 0.1197, and 0.0682, respectively, and these transitions are ππ* in nature. The transition at 249 nm is due to excitation from SOMO (singly occupied molecular orbital) (104α) → 107α; the transition at 328 nm involves (104α + 103β) → (106α + 104β) and the transition at 453 nm involves 104α (SOMO) →105α, respectively, see Figure 3. The spectra of G-OH• + 12H2O, calculated using the CAM-B3LYP-PCM/6-31++G** method, are similar to those calculated at the CAM-B3LYP-PCM/6-31G* level of theory and are shifted only by 5 – 15 nm, see Figure S5 and Table S1 in the supporting information.
Using the TD-B3LYP/6-311G**//B1B95/6-31+G** method Chatgilialoglu et al.11 calculated main transition at 333 nm having shoulders at 315 nm and 353 nm for G-OH• (adduct). As far as we know there are no experimental spectra for G(C4)-OH or G(C5)-OH radicals. The spectra reported by Candeias and Steenken9g of the reaction of OH radical with 2’-deoxyguanosine by pulse radiolysis was suggested to be that of the C4 adduct but the 620 nm absorption they report is not expected for such an adduct whose longest wavelength absorption is predicted to be 453 nm (Table S1). Chatgilialoglu et al.19 suggest the 620 nm absorption is that of G(N2-H)• which is known to have a strong absorption at this wavelength. Below we show in this work that the G•+---OH− (ion-pair) also has an absorption near this wavelength.
The spectra of ion-pair (G•+---OH− + 12H2O) computed using the CAM-B3LYP-PCM/6-31G* method is shown in Figure 4. The spectra of G•+---OH− + 12H2O shown in Figure 4 shows three intense peaks located at 267 nm, 342 nm and 580 nm and these are ππ* in nature and have oscillator strength 0.1511, 0.0441 and 0.1238. The transition at 267 nm takes place from (SOMO + 85 β) → (106α + 104 β); the second transition (94β + 98β) → 104β occurs at 342 nm and the third transition 103β → 104β occurs at 580 nm. the CAM-B3LYP-PCM/6-31++G** calculation for G•+---OH− + 12H2O shows three peaks located at 272 nm, 370 nm and 665 nm with oscillator strengths 0.1511, 0.0369, 0.1123, see Figure S6 and Table S1 in the supporting information. The longest wavelength transition is clearly more affected by the inclusion of extended basis set and red shifts by 85 nm. The ion-pair (G•+---OH−) clearly results in a strong absorption at longer wavelengths in comparison to G•+ alone, see Table S1 in the supporting information.9a,20a
Figure 4.
TD-CAM-B3LYP-PCM/6-31G*//B3LYP/6-31G* calculated excited states of G•+---OH− + 12H2O. MOs dominantly involved in the transitions are also shown.
The spectra of G(N1-H)• + 13H2O shows three intense ππ* transitions located at 273 nm, 339 nm, 445 nm and a weak ππ* transition at 570 nm, and the corresponding oscillator strengths are 0.0827, 0.0912, 0.0770 and 0.0176, respectively, see Figure 5. Our calculated transitions are in very good agreement with the absorption spectra of G(N1-H)• recorded between 300 nm – 700 nm from the pulse radiolysis at ambient temperature19 and also UV-vis absorption spectrum of G(N1-H)• in dGuo in glassy system recorded at 77 K.20 The vertical transition energies of oxidized 9-methylguanine (N1-deprotonated 9-methylguanine) calculated using the B3LYP/6-31G* method predicted transitions at 290, 350 and 470 nm.19 The spectra of G(N1-H)• + 13H2O, calculated using the CAM-B3LYP-PCM/6-31++G** method, are similar to those calculated at the CAM-B3LYP-PCM/6-31G* level of theory, see Figures 5 and S7 and Table S1 in the supporting information.
Transient absorption spectra of the reaction of OH radical with 2’-deoxyguanosine in the 300 nm – 700 nm range were reported from pulse radiolysis experiments.9g,11 The spectra recorded after 1µs of the pulse shows absorption at 300 nm – 310 nm and at ca. 400 nm and an absorption at 620 nm which decays in time. A similar spectral feature was also observed with the other guanine derivatives (guanosine, guanosine-5'-phosphate, 2'- deoxyguanosine-5'-phosphate, 1-methylguanosine and 9-methylguanine).9g,11 The absorption at 620 nm is very close to our calculated spectra of ion-pair showing transition at 580 nm and 665 nm using the CAM-B3LYP-PCM/6-31G* and CAM-B3LYP-PCM/6-31++G** methods. However, the absorption spectra of G(N2-H) radical, obtained from the pulse radiolysis at ambient temperature and in the glassy system at 77 K, also absorbs strongly at 610 nm.19, 20b Since the lifetime of the ion-pair is expected to be short, it is most likely as proposed by Chatgilialoglu et al, that the 610 nm absorption observed at the microsecond timescale is associated with the G(N2-H) radical. However, the C4/C5 pathway is clearly expected to account for a significant fraction OH radical attack and proceed via the ion-pair pathway to the G(N1-H) radical. Thus the relative yields of the G(N1-H) radical vs the G(N2-H) radical at the sub microsecond time scale would give an estimate of fraction of each of the pathways. Along this line we note that the reaction of OH radical with the model compound aniline shows that the division between the abstraction and addition/elimination pathways. Direct abstraction from anilines exocyclic NH2 accounts for 36% of the OH radicals, ca. 54% are accounted for by addition of OH radical to the ortho position of aniline followed by H2O elimination and ca. 10% irreversible addition of OH-radical to the para position of aniline is also reported.11,25 The addition reaction of OH radical with benzene, idobenzenes and other aromatic organics in aqueous solutions show the formation of radical cation through water elimination.26 These previous results strongly suggest that for OH• attack on guanine both abstraction and addition elimination mechanisms are in effect.
Summary and Conclusion
The present DFT investigation of OH radical reaction with guanine in aqueous environment confirms that OH• can react with guanine through adduct formation at C4, C5 or C8 atom or by direct abstraction of a hydrogen atom from N1 or N2 sites. Our calculations show that the thermodynamic energy of initial binding of OH• with guanine in a hydrated environment (or in the gas phase) is in this order C8OH•> C4OH• > C5OH• > (H-bonded to N1). The binding is found to be exothermic in nature and lie in the range ca. −35 kcal/mol to −15 kcal/mol, see Figure 1. Since the C8 adduct and its chemistry are well understood,2 – 5, 9 we focused in this work on the relation between C4/C5 addition and N1/N2 abstraction pathways. Under the influence of the hydration the initially formed OH adducts at C4 or C5 that have no barrier for their formation on further reaction have only a small activation barrier of ca. 1 – 6 kcal/mol to the formation of the more stable ion-pair (G•+---OH−). The resulting ion-pair (G•+---OH−) undergoes proton transfer reaction either from N1 or N2 site of guanine to form neutral radical, i.e., (G(N1-H)• or G(N2-H)•) and H2O. The present study shows that N1 deprotonation is favored over N2 deprotonation because of lower barrier height, see Figure 2. Our CAM-B3LYP calculations predict absorption between 580 nm and 670 nm for ion-pair which brackets the 620 nm absorption spectra found in pulse radiolysis studies of the OH radical reaction with 2’-deoxyguanosine.8a,9g,11 However, we note that the lifetime of such an ion-pair transient would be expected to be quite short as predicted by Mundy and coworkers13 from MD simulations. Thus, considering the relatively low barrier inherent to OH abstraction reactions it is clear that there must be a contribution to radical production from OH radical direct hydrogen abstraction from the NH2 group as proposed by Chatgilialoglu et al.11 However, the present calculations show that in the presence of hydration the barrier for direct hydrogen abstraction from N1 or N2 site may be significant (see Figure S9), whereas, for addition barrierless formation of the addition complex is predicted which then proceeds with a small barrier to form the ion-pair and products (Figure 2). As pointed out above, the reaction of OH radical with similar structures such as aniline shows a predominance of the addition/elimination reaction over the abstraction reaction from the NH2 group.11,25,26 Thus, for OH attack on guanine from both our and other theoretical treatments10,13 and experiments on similar structures, it would appear that addition/elimination should be a very competitive route with direct H-atom abstraction. Further, G(N1-H)• and G(N2-H)• can be products of both addition/elimination and abstraction which makes the distinction of the pathways of formation of G(N1-H) and G(N2-H) radicals quite difficult unless the C4/C5 adducts or ion-pair species are observed by experiments at short times. Thus experiments which look at shorter time scales are strongly encouraged. In dsDNA the picture would, of course, be altered by base stacking, base pairing and solvent accessibility.7, 14 Stacking would hinder addition reactions by lowering the accessibility and H-bonding between the DNA bases would inhibit H-abstraction reactions. These effects need to be investigated in future theoretical studies. Our study, also, shows that the CAM-B3LYP functional in conjunction with B3LYP optimized geometries provide transition energies that favorably compare to experiment for the radical species considered in this work (Table S1).
Supplementary Material
Acknowledgment
This work was supported by the NIH NCI under Grant No. R01CA045424, and computational studies were supported by a computational facilities Grant, NSF CHE-0722689. The authors thank Dr. A. Adhikary for many helpful discussions.
Footnotes
Supporting Information
B3LYP/6-31G* optimized geometries of C4, C5 (adducts), ion-pair, G(N1-H)• and G(N2-H)• due to reaction of guanine and OH radical in the presence of explicit water molecules and their spin density plots. Excited state spectra of hydrated adduct, ion-pair and G(N1-H)• calculated using the CAM-B3LYP-PCM/6-31++G** method. PES for direct hydrogen abstraction from N1 and N2 of guanine by OH radical in the presence of water molecules.
References
- 1.Becker D, Sevilla MD. In: Advances in Radiation Biology. Lett J, editor. Vol.17. Academic Press; 1993. pp. 121–180. [Google Scholar]
- 2.von Sonntag C. Free-radical-induced DNA Damage and Its Repair. Berlin, Heidelberg: Springer-Verlag; 2006. pp. 335–447. [Google Scholar]
- 3.Burrows CJ, Muller JG. Chem. Rev. 1998;98:1109. doi: 10.1021/cr960421s. [DOI] [PubMed] [Google Scholar]
- 4.(a) Cadet J, Douki T, Ravanat JL. Acc. Chem. Res. 2008;41:1075. doi: 10.1021/ar700245e. [DOI] [PubMed] [Google Scholar]; (b) Berger M, Cadet JZ. Naturforsch. B. 1985;40:1519–1531. [Google Scholar]; (c) Shukla LI, Adhikary A, Pazdro R, Becker D, Sevilla MD. Nucleic Acids Res. 2004;32:6565–6574. doi: 10.1093/nar/gkh989. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Steenken S. Chem. Rev. 1989;89:503. [Google Scholar]
- 6.(a) Van Hemmen JJ. Int. J. Radiat. Biol. 1975;27:403. doi: 10.1080/09553007514550381. [DOI] [PubMed] [Google Scholar]; (b) Scholes G. In: Radiation Chemistry of Aqueous Systems. Stein G, editor. New York: Interscience; 1968. p. 259. [Google Scholar]; (c) Scholes G. In: Photochemistry and Photobiology of Nucleic Acids. Wang SY, editor. Vol. 1. New York: Academic; 1979. p. 521. [Google Scholar]; (d) Sevilla MD. In: Excited States in Organic Chemistry and Biology. Pullman B, Goldblum N, editors. Dordrecht: Reidel; 1977. p. 15. [Google Scholar]; (e) Box HC. Radiation Effects, ESR and ENDOR Analysis. New York: Academic; 1977. [Google Scholar]; (f) Myers LS. In: Free Radicals in Biology. Pryor WA, editor. Vol. 4. New York: Academic; 1980. p. 95. [Google Scholar]; (g) Bernhard WA. Adv. Radiat. Biol. 1981;9:199. [Google Scholar]; (h) Cadet J, Berger M. Int. J. Radiat. Biol. Relat. Stud. Phys. Chem. Med. 1985;47:127. doi: 10.1080/09553008514550201. [DOI] [PubMed] [Google Scholar]; (i) von Sonntag C, Schuchmann HP. Int. J. Radiat. Biol. Relat. Stud. Phys. Chem. Med. 1986;49:1. doi: 10.1080/09553008514552201. [DOI] [PubMed] [Google Scholar]; (j) Close DM. Mag. Reson. Rev. 1988;14:1. [Google Scholar]
- 7.(a) Pogozelski WK, Tullius TD. Chem. Rev. 1998;98:1089. doi: 10.1021/cr960437i. [DOI] [PubMed] [Google Scholar]; (b) Balasubramanian B, Pogozelski WK, Tullius TD. 1998 Proc. Natl. Acad. Sci. USA. 1998;95:9738. doi: 10.1073/pnas.95.17.9738. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.(a) O'Neill P. Radiat. Res. 1983;96:198. [PubMed] [Google Scholar]; (b) O'Neill P. In: Life Chemistry Reports Supplement, Oxidative Damage and Related Enzymes. Rotilio G, Bannister JV, editors. London: Harwood Academic; 1984. p. 337. [Google Scholar]; (c) O'Neill P, Chapman PW. Int. J. Radiat. Biol. Relat. Stud. Phys. Chem. Med. 1985;47:71. doi: 10.1080/09553008514550101. [DOI] [PubMed] [Google Scholar]; (d) O'Neill P, Chapman PW, Papworth DG. Life Chem. Rep. 1985;3:62. [Google Scholar]; (e) Singh TA, Rao BSM, O'Neill P. J. Phys. Chem. B. 2010;114:16611. doi: 10.1021/jp1070049. [DOI] [PubMed] [Google Scholar]
- 9.(a) Candeias LP, Steenken S. J. Am. Chem. Soc. 1989;111:1094. [Google Scholar]; (b) Candeias LP, Wolf P, O'Neill P, Steenken S. J. Phys. Chem. 1992;96:10302. [Google Scholar]; (c) Vieira AJSC, Steenken S. J. Phys. Chem. 1987;91:4138. [Google Scholar]; (d) Vieira AJSC, Steenken S. J. Am. Chem. Soc. 1987;109:7441. [Google Scholar]; (e) Vieira AJSC, Steenken S. J. Am. Chem. Soc. 1990;112:6986. [Google Scholar]; (f) Vieira AJSC, Steenken S. J. Phys. Chem. 1991;95:9340. [Google Scholar]; (g) Candeias LP, Steenken S. Chem. Eur. J. 2000;6:475. doi: 10.1002/(sici)1521-3765(20000204)6:3<475::aid-chem475>3.0.co;2-e. [DOI] [PubMed] [Google Scholar]
- 10.(a) Colson AO, Sevilla MD. J. Phys. Chem. 1995;99:3867. [Google Scholar]; (b) Colson AO, Sevilla MD. J. Phys Chem. 1996;100:4420. [Google Scholar]; (c) Colson AO, Sevilla MD. J. Phys. Chem. 1995;99:13033. [Google Scholar]; (d) Wetmore SD, Boyd RJ, Eriksson LA. J. Phys. Chem. B. 1998;102:9332. [Google Scholar]; (e) Llano J, Eriksson LA. Phys. Chem. Chem. Phys. 2004;6:4707. [Google Scholar]; (f) Jena NR, Mishra PC. J. Phys. Chem. B. 2005;109:14205. doi: 10.1021/jp050646j. [DOI] [PubMed] [Google Scholar]; (g) Jena NR, Mishra PC. Chem. Phys. Lett. 2006;422:417. [Google Scholar]; (h) Agnihotri N, Mishra PC. Chem. Phys. Lett. 2011;503:305. [Google Scholar]; (i) Naumov S, von Sonntag C. Radiat. Res. 2008;169:355. doi: 10.1667/RR1081.1. [DOI] [PubMed] [Google Scholar]; (j) Scheiner S. Int. J. Quant. Chem. 2011 DOI: 10.1002/qua.23089. [Google Scholar]
- 11.(a) Chatgilialoglu C, D'Angelantonio M, Guerra M, Kaloudis P, Mulazzani QG. Angew. Chem. Int. Ed. 2009;48:2214. doi: 10.1002/anie.200805372. [DOI] [PubMed] [Google Scholar]; (b) Chatgilialoglu C, D’Angelantonio M, Kciuk G, Bobrowski K. Chem. Res. Toxicol. 2011 doi: 10.1021/tx2003245. DOI:10.1021/tx2003245. [DOI] [PubMed] [Google Scholar]
- 12.Cheng Q, Gu J, Compaan KR, Schaefer HF. Chem. Eur. J. 2010;16:11848. doi: 10.1002/chem.201001236. [DOI] [PubMed] [Google Scholar]
- 13.(a) Mundy CJ, Colvin ME, Quong AA. J. Phys. Chem. A. 2002;106:10063. [Google Scholar]; (b) Wu Y, Mundy CJ, Colvin ME, Car R. J. Phys. Chem. A. 2004;108:2922. [Google Scholar]
- 14.Abolfath RM, van Duin ACT, Brabec T. J. Phys. Chem. A. 2011;115:11045. doi: 10.1021/jp204894m. [DOI] [PubMed] [Google Scholar]
- 15.(a) Dreuw A, Weisman JL, Head-Gordon M. J. Chem. Phys. 2003;119:2943. [Google Scholar]; (b) Dreuw A, Head-Gordon M. Chem. Rev. 2005;105:4009. doi: 10.1021/cr0505627. [DOI] [PubMed] [Google Scholar]
- 16.(a) Yanai T, Tew DP, Handy NC. Chem. Phys. Lett. 2004;393:51. [Google Scholar]; (b) Peach MJG, Sueur CRL, Ruud K, Guillaume M, Tozer DJ. Phys. Chem. Chem. Phys. 2009;11:4465. doi: 10.1039/b822941d. [DOI] [PubMed] [Google Scholar]
- 17.Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, Scalmani G, Barone V, Mennucci B, Petersson GA, Nakatsuji H, Caricato M, Li X, Hratchian HP, Izmaylov AF, Bloino J, Zheng G, Sonnenberg JL, Hada M, Ehara M, Toyota K, Fukuda R, Hasegawa J, Ishida M, Nakajima T, Honda Y, Kitao O, Nakai H, Vreven T, Montgomery JA, Jr, Peralta JE, Ogliaro F, Bearpark M, Heyd JJ, Brothers E, Kudin KN, Staroverov VN, Kobayashi R, Normand J, Raghavachari K, Rendell A, Burant JC, Iyengar SS, Tomasi J, Cossi M, Rega N, Millam JM, Klene M, Knox JE, Cross JB, Bakken V, Adamo C, Jaramillo J, Gomperts R, Stratmann RE, Yazyev O, Austin AJ, Cammi R, Pomelli C, Ochterski JW, Martin RL, Morokuma K, Zakrzewski VG, Voth GA, Salvador P, Dannenberg JJ, Dapprich S, Daniels AD, Farkas O, Foresman JB, Ortiz JV, Cioslowski J, Fox DJ. Gaussian 09. Wallingford CT: Gaussian, Inc.; 2009. [Google Scholar]
- 18.(a) Gauss View. Pittsburgh, PA: Gaussian, Inc.; 2003. [Google Scholar]; (b) Allouche AR. Gabedit is a free Graphical User Interface for computational chemistry packages. doi: 10.1002/jcc.21600. It is available from http://gabedit.sourceforge.net/. [DOI] [PubMed]
- 19.Chatgilialoglu C, Caminal C, Guerra M, Mulazzani QG. Angew. Chem. Int. Ed. 2005;44:6030. doi: 10.1002/anie.200501087. [DOI] [PubMed] [Google Scholar]
- 20.(a) Adhikary A, Kumar A, Becker D, Sevilla MD. J. Phys. Chem. B. 2006;110:24171. doi: 10.1021/jp064361y. and references therein. [DOI] [PMC free article] [PubMed] [Google Scholar]; (b) Adhikary A, Kumar A, Munafo SA, Khanduri D, Sevilla MD. Phys. Chem. Chem. Phys. 2010;12:5353. doi: 10.1039/b925496j. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Tripathi GNR. J. Am. Chem. Soc. 1998;120:4161. [Google Scholar]
- 22.Mitroka S, Zimmeck S, Troya D, Tanko JM. J. Am. Chem. Soc. 2010;132:2907. doi: 10.1021/ja903856t. [DOI] [PubMed] [Google Scholar]
- 23.Kumar A, Sevilla MD. J. Phys. Chem. B. 2009;113:11359. doi: 10.1021/jp903403d. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Adhikary A, Khanduri D, Sevilla MD. J. Am. Chem. Soc. 2009;131:8614. doi: 10.1021/ja9014869. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.(a) Solar S, Solar W, Getoff N. Radiat. Phys. Chem. 1986;28:229. [Google Scholar]; (b) Christensen H. Int. J. Radiat. Phys. Chem. 1972;4:311. [Google Scholar]
- 26.(a) Mohan H. Radiat. Phys. Chem. 1997;49:15. [Google Scholar]; (b) Eberhardt MK. J. Am. Chem. Soc. 1981;103:3876. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.