Deoxyriboadenosine (dA) − Deoxyribothymidine - American Chemical

J. Phys. Chem. B 2005, 109, 13067-13075

13067

FEATURE ARTICLE Structural and Energetic Characterization of a DNA Nucleoside Pair and Its Anion: Deoxyriboadenosine (dA) - Deoxyribothymidine (dT) Jiande Gu,*,†,‡ Yaoming Xie,‡ and Henry F. Schaefer, III*,‡ Drug Design & DiscoVery Center, State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica, Shanghai Institutes for Biological Sciences, CAS, Shanghai 201203 P. R. China, and Center for Computational Chemistry, UniVersity of Georgia, Athens, Georgia 30602-2525 ReceiVed: March 27, 2005; In Final Form: May 1, 2005

The geometries of the DNA nucleoside pairs between 2′-deoxyriboadenosine (dA) and 2′-deoxyribothymidine (dT) and its anion (dAdT-) were fully optimized using carefully calibrated density functional methods. The addition of an electron to dAdT results in remarkable changes to the two hydrogen bonding distances, the H‚‚‚O distance decreasing by 0.303 Å and the N‚‚‚H distance increasing by 0.229 Å. The electron affinity of the dAdT pair was studied to reveal the correct trends of adiabatic electron affinity (EAad) under the influence of the additional components to the individual bases. The consequence of negative charge in terms of structural variations, energetic changes, and charge distribution were explored. The EAad of dAdT is predicted to be positive (0.60 eV), and it exhibits a substantial increase compared with those of the corresponding bases A and T and the nucleic acid base pair AT. The effects of pairing and the addition of the sugar moiety on the EAad are well described as the summation of the individual influences. The influence of the pairing on the EA is comparable to that of the addition of 2-deoxyribose. The excess charge is mainly located on the thyminyl moiety in the anionic dAdT pair. The positive vertical electron affinity (VEA ) 0.20 eV) for dAdT suggests that it is able to form a stable anion through electron attachment. A large vertical detachment energy (VDE ) 1.14 eV) has been determined for the anionic dAdT nucleoside pair. Therefore, one may expect that the stable anionic dAdT nucleoside pair should be able to undergo the subsequent glycosidic bond cleavage process.

I. Introduction Several important processes in biochemistry are closely related to the formation of ions in DNA fragments; electron trapping within nucleobase sites has been thought to play a crucial role in DNA damage and repair.1-6 Charge transfer along DNA could be impaired by the formation of stable anions of nucleosides.7-15 Acquisition of excess charge is likely responsible for the cascade of reactions leading to mutation.1,3,4,16 Recently, experimental and theoretical studies have demonstrated that, even at very low energies, electrons may induce strand breaks in DNA via dissociative electron attachment.17-22 Therefore, the knowledge of the distribution of excess electron sites and of reliable electron affinities for DNA fragments is of great importance. Efforts toward understanding the effects of excess electrons on DNA have been focused on the determination of electron affinities extending from individual nucleic acid bases (NAB) to base pairs, and from nucleobases to nucleosides to nucleotides. Although the individual NABs in the gas phase are in principle experimentally observable, the precise determination of their electron affinities remains challenging.4,23-30 Early experimental estimates of electron affinities suggested that all * Corresponding authors. E-mail: [email protected]; [email protected]. † Shanghai Institutes for Biological Sciences. ‡ University of Georgia.

the nucleobases have substantial (0.56 ∼ 1.51 eV) adiabatic electron affinities (EAad), with the relative order being C < T < U < A , G.23,27 However, later extrapolated electron affinities derived from the microsolvation experiments are much smaller (about 0.1 eV for T, C, and U).26 Negative EA values have also been determined for A and C in gas-phase experiments.4,28 The identification of the existence of two types of anions (dipole bound and covalent anions)24,29,30 can only partly explain the controversies in the experimental EA values.27 It can be safely stated that the reliable experimental electron affinities of G, C, A, T, and U all fall in the range 0.0 ( 0.2 eV. Further progress in the investigation of the excess charge in DNA extends to the Watson-Crick base pairs in the gas phase. Some of the gas-phase isolated base pairs, including the AT anion, have been detected by mass spectrometric methods,31 and the GC pair has been identified by its resonance-enhanced multiphoton ionization spectrum.32 Gas-phase experimental studies of the isolated nucleosides have also been reported, including resonant two-photon ionization spectra of the laser desorbed, jet-cooled species cytosine, guanosine, 2′-deoxyriboguanosine, and 3′-deoxyriboguanosine;33-35 as well as FTIR studies36 of adenosine, cytidine, and uridine. A gas phase study of H/D exchange was also carried out for nucleotides.37 However, the experimental determination of the EAs for the base pairs, nucleosides, and nucleotides has not been reported.

10.1021/jp0515535 CCC: $30.25 © 2005 American Chemical Society Published on Web 06/16/2005

13068 J. Phys. Chem. B, Vol. 109, No. 27, 2005

Henry F. Schaefer, III received his S.B. degree in Chemical Physics from M.I.T. in 1966 and Ph.D. from Stanford University in 1969. The same year he joined the faculty of the University of California at Berkeley, where he was a professor for 18 years. He has received four awards from the American Chemical Society (Pure Chemistry, Baekeland, Remsen, and Theoretical Chemistry), plus the Centenary Medal from London’s Royal Society of Chemistry. He is a Fellow of the American Academy of Arts and Sciences and will receive the 20052006 Joseph O. Hirschfelder Prize of the University of Wisconsin.

Large DNA fragments, such as nucleosides and nucleotides, are nonvolatile and the requirement of the vaporization of the species without thermal degradation makes the observation in the gas-phase challenging. Theoretical investigations at various levels of sophistication have complemented the above experiments. Theoretical predictions of the individual nucleobase EAs have produced a range of results.30,38-45 While second-order Møller-Plesset perturbation theory (MP2) with a modest basis set including diffuse functions yields negative EAad values ranging from -1.19 to -0.25 eV for the five bases,41,45 density functional theory (DFT) methods with good basis sets predict positive EAad values for both U and T.41,43,44 The accurate determination of these electron affinities is thus challenging for both experiment and theory. However, recent synergy between the two has resulted in the development of a DFT bracketing technique that has advanced the understanding of the interaction of molecules with excess negative charge by providing a reliable method for electron affinity determinations.46 For the 91 molecules for which reliable ((0.1 eV) experimental electron affinities exist, the average absolute error with the B3LYP method is 0.14 eV. With this B3LYP approach, experiment-consistent EAad values close to zero ((0.25 eV) are predicted for all five members of this important family of molecules, with the ordering U > T > C ∼ G > A.43 Theoretical studies of excess charge in DNA have extended the prediction of EAs to the Watson-Crick base pairs. Based on the positive EA values determined by photoelectron spectra for the thymine-water and cytosine-water complexes,26 it is expected that the EAad values for the GC and AT pairs are also positive. However, negative EAad values for the GC pair and the AT pair have been predicted by HF and MP2 studies.47-49 For instance, the MP2 method estimated the EA for the AT pair to be more negative than -0.4 eV.48 MP2 theory also predicted a negative EAad for the microsolvated model uracil‚ (H2O)3,50,51 despite the anion stabilization effects of microsolvation observed in the laboratory.26,28 Even with a much larger basis set (6-311++G(2df,2p)), MP2 still underestimates EAad for uracil‚H2O by ∼50%.52 On the other hand, the DFT studies achieved reasonable descriptions of the EAs of the WatsonCrick base pairs.44,53-57 From the best estimates, the influence of G on C increases the EA from C to the GC pair by an amount

Gu et al.

Jiande Gu received his B.S. and Ph.D. (1995) degrees in chemistry from Sichuan University, Chengdu, China. Between 1998 and 2000 he was a postdoctoral research fellow with Professor F. Albert Cotton at Texas A&M University. He has been a professor of chemistry at the Shanghai Institute of Materia Medica, Chinese Academy of Sciences, since 1998. Dr. Gu is a frequent visiting professor at the Center for Computational Chemistry, University of Georgia, and the Computational Center for Molecular Structure and Interactions, Jackson State University. His primary research interests lie in the area of theoretical biochemistry.

Yaoming Xie received his M.S. degree in physical chemistry from Sichuan University (Chengdu, China) and his Ph.D. from the University of Texas (Austin) in 1988. He is currently a senior research associate at the Center for Computational Chemistry, University of Georgia. Dr. Xie is also a guest professor at Sichuan University and an adjunct professor at the Beijing Institute of Technology. His research interests range from theoretical organometallic chemistry to biochemistry.

equivalent to that of two water molecules, and that of A on T by about one water molecule.57 For the nucleosides and nucleotides, semiempirical methods have been applied to evaluate EAs.58,59 Recently, with reliably calibrated B3LYP/DZP++ approaches, a more accurate bracketing of the electron affinities of the 2′-deoxyribonucleosides has been accomplished.60 Due to the influence of the 2′deoxyribose, the EAad values for dA, dG, dC, and dT are all positive. The presence of 2′-deoxyribose at atom N1 of the primidines or at N9 of the purines increases the EAad values by 0.16 ∼ 0.34 eV (compared to the individual bases) in a similar way as microsolvation.60 These findings raise new questions as to the effects of additional components to the system. In the microsolvation experiments, the EAs are found to be almost linearly proportional to the number of solvating water molecules.26 It is of interest to examine whether this additive phenomenon might also be seen for the nucleoside pairs. Of particular interest will be the residual charge distribution in the nucleoside pairs. Extension from the individual nucleosides to

Feature Article SCHEME 1

J. Phys. Chem. B, Vol. 109, No. 27, 2005 13069 TABLE 1: Electron Affinities of Nucleic Acid Bases, Nucleosides, Base Pairs, and Nucleoside Pairs (in eV) A T dA dT AT dAdT

the nucleoside pairs should provide answers to the above questions and shed light on the physical background for the preferential location of the excess charge. Also, the use of the calibrated DFT approach should yield reliable estimates for the adiabatic EAs of the nucleoside pairs, and the extension from the nucleobase pairs to the nucleoside pairs is a crucial step toward the understanding the delicate relationship between the structure and function of DNA. In this research, we have predicted the adiabatic electron affinity of the 2′-deoxyriboadenosine (dA) 2′-deoxyribothymidine (dT) nucleoside pair at a reliable level of theory (B3LYP/ DZP++) in an effort to reveal the correct trends of EAs under the influence of the additional components to the individual bases. The consequences of negative charge in terms of structural variations, energetic changes, and charge distribution are also explored. The greatly tightened O4(T)‚‚‚H6(A) hydrogen bond and the elongated N6-H6 bond in the AT pair anion57 seem to suggest that proton transfer might occur between O4(T) and N6(A) and result in an additional conformer of the AT pair anion. However, experimental evidence8,61 and previous theoretical studies on the AT base pair62 suggest that the double proton transferred isomer is unfavorable in terms of energy. Therefore, our research is focused on the canonical WatsonCrick dAdT nucleoside pair. Scheme 1 provides the standard numbering of atoms in dAdT and its anion.

EAada

VEAb

VDEc

-0.37 (-0.28)d 0.06 (0.20)d -0.05 (0.06)e 0.31 (0.44)e 0.19 (0.36)f 0.43 (0.60)

-0.45 -0.30 -0.24 0.05 -0.03 0.20

-0.29 0.36 0.91e 0.94e 0.63 1.14

a Values with zero point correction are given in parentheses. b VEA ) E(neutral) - E(anion), the energies are calculated based on the optimized neutral structures. c VDE ) E(neutral) - E(anion), the energies are calculated based on the optimized anion structures. d Reference 43. e Reference 60. f Reference 57.

centered on each heavy atom. The even-tempered orbital exponents were determined according to the prescription of Lee and Schaefer:68

Rdiffuse )

(

)

1 R1 R2 + R 2 R2 R3 1

where R1, R2, and R3 are the three smallest Gaussian orbital exponents of the s- or p-type primitive functions for a given atom (R1 < R2 < R3). The final DZP++ set contains six functions per H atom (5s1p/3s1p) and nineteen functions per C, N, or O atom (10s6p1d/5s3p1d). For the dA-dT complex there are a total of 62 atoms, 180 degrees of freedom, and 827 basis functions. This basis has the tactical advantage that it has previously been used in comprehensive calibrative studies46 of a wide range of electron affinities. To analyze the distribution of the unpaired electron, (KohnSham) molecular orbitals and spin density plots were constructed from the appropriate B3LYP/DZP++ results. Charge distributions were characterized through the electrostatic potential maps. Natural population analysis (NPA) charges were determined using the B3LYP functional and the DZP++ basis set with the natural bond order (NBO) analysis of Reed and Weinhold.69-72 III. Results and Discussion

II. Theoretical Methods Previous research has demonstrated that the B3LYP method predicts reasonable results for DNA bases, base pairs, and anions.43,57,60 The B3LYP functional is a combination of exchange from Becke’s 3-parameter HF/DFT hybrid exchange functional (B3)63 with the dynamical correlation functional of Lee, Yang, and Parr (LYP).64 In this study, optimized geometries, absolute vibrationally zero-point corrected energies, and natural charges for the nucleoside pair formed by 2′-deoxyriboadenosine (dA) and 2′-deoxyribothymidine (dT) were determined using the B3LYP approach. For the neutral complex, the restricted B3LYP formalism is employed, while the unrestricted formalism is used for anion. The GAUSSIAN 98 systems of DFT programs65 was used for all computations. The B3LYP method is adopted along with double-ζ quality basis sets with polarization and diffuse functions (denoted as DZP++). The DZP++ basis sets were constructed by augmenting the Huzinage-Dunning66,67 set of contracted double-ζ Gaussian functions with one set of p-type polarization functions for each H atom and one set of five d-type polarization functions for each C, N, and O atom (Rp(H) ) 0.75, Rd(C) ) 0.75, Rd(N) ) 0.80, Rd(O) ) 0.85). To complete the DZP++ basis, one even-tempered diffuse s function was added to each H atom while sets of even-tempered diffuse s and p functions were

A. Electron Affinities. The adiabatic electron affinity was predicted as the difference between the total energies of the appropriate neutral and anion at their respective optimized geometries

EAad ) Eneut - Eanion The adiabatic electron affinity of the nucleoside pair dAdT exhibits a substantial increase as compared to that of the corresponding nucleic acid base pair AT (0.60 eV vs 0.36 eV, see Table 1). The increase of the EAad from the AT pair to the dAdT pair amounts to 0.24 eV, demonstrating that the addition of 2-deoxyribose to the bases increases the electron affinity for the combined moiety over the corresponding nucleic acid base pair. This effect is consistent with that predicted for the isolated nucleoside thymidine in a previous study,60 in which the EAad increase from T to dT is about 0.24 eV. Since both A and dA have much smaller EAad value than those of T and dT, one may expect that the excess electron resides primarily on the thymidine moiety within the anionic nucleoside dAdT- pair, resembling the situation for the AT- base pair anion.57 This similarity may be further justified by the examination of the molecular orbital which the “last” electron occupies. Plots of the singly occupied molecular orbitals (SOMOs) for the related

13070 J. Phys. Chem. B, Vol. 109, No. 27, 2005

Gu et al.

Figure 1. Plots of the singly occupied molecular orbitals (SOMOs) for the nucleic acid base anions and the base pair anion as well as the corresponding nucleoside anions.

Figure 2. Significant changes (>0.01 au) in the natural population analysis (NPA) charges of the dAdT pair that occur between the neutral and anionic species (in au).

nucleic acid base anions and the base pair anion AT-, as well as the corresponding nucleoside anions, are shown in Figure 1. The SOMO of the dAdT- anion is seen to display the characteristics of the SOMO of the dT- anion; similarly, the SOMO of the AT- anion pair exhibits the main features of the anion of the base T. The influence of nucleoside pairing on the adiabatic electron affinity raises the EAad of isolated thymidine by 0.16 eV (0.44 eV for dT and 0.60 eV for dAdT). For comparison, the analogous pairing increases the EAad of the base thymine by the same amount, 0.16 eV (0.20 eV for T and 0.36 eV for AT). The effect on the EAad due to pairing and the addition of the sugar moieties is equal to the summation of the individual effects. Examining the SOMOs of the AT- or dAdT- pair anions reveals that the excess electron is primarily located on the T or dT parts of these molecular systems. Delocalization of the excess electron over the base pair seems to further stabilize the anions. The location of charge on the constituent parts of the nucleoside pair also provides some insight into the overall electronic effects of negative charge. Figure 2 illustrates the significant changes (>0.01 au) in the dAdT pair atomic charges that occur between

TABLE 2: Natural Population Analysis (NPA) Charges Residing on the 2-Deoxyribose and on the Base within the Anionic Thymidine, Anionic DAdT Pair, and Anionic AT Pair dAdTATdTa

A

T

Ribose

-0.07 -0.08a

-0.81 -0.92a -0.87

-0.12 s -0.13

Reference 57.

the neutral and anionic species. The excess charge is again shown to be mainly on the thyminyl moiety of dT. Table 2 summarizes the extra charge distribution among 2-deoxyribose and the bases in the anionic complex. In the anionic nucleoside pair, the total negative charge increases on the A, T, and 2-deoxyribose (linked to T) parts amount to 0.07, 0.81, and 0.12 au, respectively, compared to neutral dAdT pair. No charge alteration has been detected for the sugar bonded to adenine. The extra charge is mainly located on the thyminyl group. The analysis of the NPA charge difference between the neutral and anionic AT pair shows that there is 0.92 of negative charge located on T and 0.08 au on A.57 On the other hand, the charge distribution difference of the isolated dT shows that 0.87 au of

Feature Article

J. Phys. Chem. B, Vol. 109, No. 27, 2005 13071

Figure 3. SOMOs of the vertically electron attached anions. These anions have partial dipole-bound character at their primitive stage of formation.

the extra charge resides on the thyminyl group, and 0.13 is on the 2-deoxyribose.60 The distribution of the excess electron among the constituent parts seems directly correlated to the base EAad values. Generally, the influence of the pairing on the electron affinity of the base is less than that of the addition of 2-deoxyribose. The vertical electron affinity (VEA), which determines the necessary energy needed in a fast electron capture step, is an important physical property to understand the formation of anions.73 The positive VEA values of dT and dAdT suggest that both are good electron captors. The fact that the VEA of dT amounts to 0.05 eV and the larger VEA value of 0.20 eV for the dAdT nucleoside pair indicates that the hydrogenbonding between dA and dT significantly improves the ability of AT to attract an electron. The SOMOs of the vertically electron attached anions depicted in Figure 3 reveal that these anions (especially A-) may have some dipole-bound characteristics at this primitive stage of formation. We note that a dipole-bound electron in anions is less stable and easily detached (EA typically less than 0.1 eV). Only those anions with positive VEA values can have lifetimes long enough to transform into stable covalently bound anions through structural relaxation. According to the present theoretical estimations of the VEA and EAad values, the dAdT nucleoside pair should be able to form a stable anion through electron attachment. To examine the lifetime of the stable dAdT- pair anion, the vertical detachment energy has also been evaluated. A substantial vertical detachment energy (VDE) is predicted, 1.14 eV based on the optimized anion structure of the dAdT- nucleoside pair. For comparison, the VDEs of dT- and AT- are 0.94 and 0.63 eV, respectively. Pairing between dA and dT increases the VDE of the dT- anion by about 0.2 eV. Recent experimental and theoretical studies reveal that the extra negative charge in nucleoside and nucleotide anions might induce glycosidic bond breaking in nucleosides74-76 or phosphate-sugar C-O σ bond cleavage in nucleotides.17,73,77-81 With a VDE value of 1.14 eV, one may expect that once the stable anionic dAdT- nucleoside pair is formed, it should have lifetime long enough to undergo the subsequent glycosidic bond cleavage process. B. Geometries. The fully optimized geometries of the neutral and anionic 2′-deoxyriboadenosine 2′-deoxyribothymidine pairs are depicted in Figures 4 and 5, respectively. Cartesian

coordinates for all atoms in dAdT and dAdT- are provided as Supporting Information (Tables S1 and S2). Compared to the isolated nucleosides, the obvious geometric variations involve the structural features in the hydrogen bonding. Due to the H-bonding between H(N6) of dA and O4 of dT, the pyramidal structure of N6 about the isolated dA60 takes the planar form in both the neutral and anionic dAdT nucleoside pairs, similar to the structure change in the AT base pair.57 In the neutral dAdT pair, there is a slight increase in the H6-N6 bond length (1.027 Å vs 1.012 Å of dA60) and a minor elongation of O4-C4 (1.240 Å vs 1.227 Å of dT60), indicating the formation of the H6(dA)‚ ‚‚O4(dT) H-bond, similar to the case of AT from A and T. The H-bonding between N1 of dA and H(N3) of dT lengthens the H-N3 bond distance by 0.037 Å compared to dT.60 The planar arrangement of N6 in dA extends the conjugated π bond of the purine to N6, causing minor increases in the N1-C2 and N1C6 bonds (by about 0.01 Å) and a slight decrease in C6-N6 bond (about 0.02 Å) of dA. The H-bond distances between dA and dT are 1.890 Å and 1.797 Å for H6(dA)‚‚‚O4(dT) and N1(dA)‚‚‚H3(dT), respectively. It should be noted that the H-bond parameters and the overall geometric parameters of the adeninyl and thyminyl moieties in the neutral dAdT pair are extremely close to those in the AT base pair, the differences being less than 0.002 Å. Considering that the predicted dipole moment of dA (2.39 D) is close to that of A (2.51 D) and the dipole moment of dT (4.46 D) is similar to that of T (4.59 D),60 this H-bonding similarity between dAdT and AT is not unexpected. The outstanding structural discovery in the previous reseach57 on AT- was the remarkable change in hydrogen bond distances with respect to neutral AT. The same conspicuous structural change is found here for the dAdT f dAdT- transition. For neutral dAdT the H‚‚‚O hydrogen bond distance is 1.890 A, a bit less than that for the water dimer, ∼1.94 Å.82 However, for the dAdT- anion, this H‚‚‚O distance is much shorter, 1.578 Å. This decrease in H‚‚‚O distance is perhaps due to the attraction between the greater electron density of dT in dAdTand the partial positively charged hydrogen atom. The latter distance is so short as to raise some questions about possible partial chemical bonding, rather than conventional hydrogen bonding. The N‚‚‚H distance in dAdT is predicted to be 1.797 Å, while the analogous distance for the dAdT- anion is much longer, 2.026 Å. The similarity of these structural changes to those for AT f AT- suggests that this shift in hydrogen bond distances may be characteristic of DNA fragment anions.

13072 J. Phys. Chem. B, Vol. 109, No. 27, 2005

Gu et al.

Figure 4. Fully optimized geometries of the neutral 2′-deoxyriboadenosine 2′-deoxyribothymidine pair (in Å).

Figure 5. Fully optimized geometries of the anionic 2′-deoxyriboadenosine 2′-deoxyribothymidine pair (in Å).

C. Nucleoside Pairing Interaction. The molecular orbital analysis and the NPA results clearly demonstrate that the excess charge mainly resides on dT in the dAdT- pair anion. The geometrical character of the anionic dAdT- pair is consistent with this observation. The geometrical parameters of the dT portion in dAdT- show typical anionic character, while the geometry of dA in the anionic pair keeps the feature of the neutral species. The C6 atom next to the glycosidic bond in thymidine exhibits sp3 hybrid characteristics, consistent with that for the isolated thymidine anion. Due to the pairing

interaction, the elongation of the O4-C4 bond of dT amounts to 0.022 Å (from 1.263 Å to 1.285 Å), slightly longer than that for the neutral complex. On the other hand, the increase in the H3-N3 bond length of dT is less significant, only 0.014 Å (from 1.013 to 1.027 Å) longer than that for the isolated anionic thymidine. The location of the excess electron on dT is expected to increase the negative charge density near atoms O4 and N3 of dT. Accordingly, the proton accepting ability of O4 and N3 is reinforced, consistent with the longer O4-C4 bond and the less

Feature Article

J. Phys. Chem. B, Vol. 109, No. 27, 2005 13073

Figure 6. Significant geometrical changes (>0.01 Å) in the dAdT pair between the neutral and anionic species (in Å). Note in particular, the large changes in the H‚‚‚O (-0.303 Å) and N‚‚‚H (0.229 Å) hydrogen bond distances.

TABLE 3: Dissociation Energies of the Neutral and Anionic Pairs (in kcal/mol) dAdT f dA + dT dAdT- f dA + dTAT f A + T AT- f A + Tb

∆E

∆EZPVE a

-14.2 -16.9 -13.8b -16.7b

-13.0 -16.0 -12.5b -16.1b

a ∆EZPVE: Zero-point vibrationally corrected dissociation energies. Reference 57.

elongated H3-N3 bond of dT in the anionic pair mentioned above. As a result, a short hydrogen-bond length of 1.587 Å for H6(dA)‚‚‚O4(dT) and a long hydrogen-bond length of 2.026 Å for the N1(dA)‚‚‚H3(dT) are detected for the anionic dAdT pair. Comparison of the geometric alterations between the neutral and anionic dAdT nucleoside pair further supports the above rationale. Figure 6 displays the significant geometric changes (>0.01 Å) between the neutral dAdT pair and its anion. The substantially reduced H6(dA)‚‚‚O4(dT) bond (0.303 Å shorter) and the elongated N6-H bond of dA and O4-C4 bond of dT (0.041 Å and 0.045 Å longer) demonstrate the modification of the H-bonding in the paired anion. Meanwhile, the greatly increased N1(dA)‚‚‚H3(dT) atomic distance (0.229 Å longer) and the significant shortening of the H3-N3 bond in dT (0.027 Å shorter) indicate that the proton donating ability at the H3N3 site of dT is much weaker. The charge distribution difference between the anionic dAdTpair and the anionic AT- pair also has influences on the hydrogen-bond geometry. Compared to the anionic AT- pair, the H6(dA)‚‚‚O4(dT) bond of the dAdT- pair (1.587 Å) is slightly longer than that for AT- (1.573 Å),57 and the N1(dA)‚ ‚‚H3(dT) bond is shorter (2.026 Å in dAdT- vs 2.071 Å in AT-).57 These differences are in accord with the 0.81 of excess charge located on the thyminyl moiety in the dAdT- pair and 0.92 au of charge on thymine in the AT- pair.57 The hydrogen-bonding variations due to the attachment of an excess electron on the dAdT pair also influence the pairing energy. Table 3 provides the pairing energies of the neutral and anionic dAdT pairs, along with those of AT and AT- for comparison. The dissociation energy of the neutral dAdT is predicted to be 14.2 kcal/mol, about 2.7 kcal/mol smaller than that for the anionic complex. This dissociation energy increase in the anionic dAdT- pair suggests that the H-bonding energy gained in the intensified H6(dA)‚‚‚O4(dT) interaction is larger than the energy loss in the weakened N1(dA)‚‚‚H3(dT) bond. This result is reasonable, considering the fact that the H-bonding interaction of the H‚‚‚O type is stronger than that of the H‚‚‚N

variety in general. The pairing energies of the nucleoside pairs dAdT and dAdT- are found to be close (within 1 kcal/mol at the same theoretical level) to those of the nucleic acid base pairs AT and AT- (as shown in Table 3). This value is also close to the most reliable recent results (13.8-15.8 kcal/mol at various levels of theory).83 The addition of the sugar moiety seems to have little influence on the pairing energy. Similarly, the pairing energies are expected to be close to those for the nucleotide pairs with the presence of the counterions tightly bound to the phosphate groups. D. Vibrational Frequencies. Harmonic vibration analyses of both neutral and anionic dAdT nucleoside pairs reveal the influences of the addition of the sugar moiety on the interbase vibrational modes and frequencies. The complete list of the harmonic vibrational frequencies and their infrared intensities are given in Supporting Information (Table S3 and Table S4). The only pure internucleoside vibrational mode to be identified is the dA-dT out-of-plane motion, which has the lowest harmonic frequency (ω1) in both neutral (11 cm-1) and anionic (19 cm-1) pairs. As comparison, the lowest frequency for the similar motion for AT and AT- amounts to 29 and 23 cm-1, respectively.57 The other vibrational modes for the dAdT pair combine the internucleoside motions with the intranucleoside motion. These vibrational modes are expected to have lower frequencies than those for AT base pairs.57 The relative motions between the base and the ribose are found to be the main contributions to these intranucleoside modes corresponding to the frequencies ω2 - ω11. The H-bond stretching modes (mixed with sugar-ring twisting) for the neutral dAdT pair are located at 93-99 cm-1 (ω12 - ω14), which are slightly lower than that for the AT pair (ω7, 118 cm-1).57 One H-bond stretching mode for the anionic dAdT- pair is 109 cm-1 (ω14), which is very slightly lower than that in the corresponding AT- pair (ω7, 110 cm-1).57 The addition of the sugar moiety seems to decrease the interbase vibrational frequencies. The differences of the interbase vibrational frequencies between the neutral and the anionic dAdT pairs are consistent with the previous conclusion that the dAdT H-bonding is reinforced in the anionic pair, e.g., 11 cm-1 vs 19 cm-1 for dA-dT out-of-plane mode, 93-99 cm-1 vs 109 cm-1 for H-bond stretching modes, and 23 cm-1 vs 31 cm-1 for the base plane twisting mode (Table S4). It should be noted that these low-frequency vibrational modes are expected to be very anharmonic,84 so our theoretically predicted vibrational frequencies might exceed those eventually determined by experiment.

13074 J. Phys. Chem. B, Vol. 109, No. 27, 2005 IV. Conclusions This exploration of electron affinities, pairing energies, geometries, and charge distributions of the neutral and anionic 2′-deoxyribonucleoside pair dAdT has shown the following: 1. The adiabatic electron affinity of the 2′-deoxyribonucleoside pair dAdT is predicted to be positive (0.60 eV) and it exhibits a substantial increase (0.24 eV) compared to that of the corresponding nucleic acid base pair AT. The effects on the AT EAad due to pairing and the addition of sugar moieties are well described by the summation of the individual effects. Generally, the influence of the pairing on the electron affinity of the base is less than that of the addition of 2-deoxyribose. 2. The excess charge is mainly located on the thyminyl moiety in the anionic dAdT pair, which is similar to the case of the AT pair. The presence of the sugar moiety in the dAdT pair does not significantly alter the distribution of the excess charge. The distribution of the excess electron among the constituent parts is correlated to the isolated base EAad values. 3. The positive vertical electron affinity (VEA) value of 0.20 eV for dAdT suggests that dAdT pair can capture the low energy incident electrons easily. Due to this positive VEA, the electron detachment is unlikely to happen during the structural relaxation process that leads to the formation of the stable covalently bound anion radical of dAdT pair. According to this theoretical estimation of VEA and EAad, the dAdT pair should be able to form a stable anion through electron attachment. 4. A large vertical detachment energy (VDE ) 1.14 eV) is predicted for the anionic dAdT nucleoside pair. Considering the relatively lower activation energy barrier for the glycosidic bond breaking predicted at the same level of theory,75 one may expect that the stable anionic dAdT nucleoside pair should be able to undergo subsequent glycosidic bond cleavage process. 5. The addition of the sugar moiety seems to have only a small (0.4 kcal/mol) influence on the pairing energy. Accordingly, the pairing energies may be close to those of the base pairs, even for the nucleotides, with the presence of the tightly bonded counterions to the phosphate groups. These findings suggest that the neutral pyrimidines in the duplex structure of DNA are good electron captors. The baserelease related DNA damage due to low energy electron attachment processes should be attributed to the substantial electron affinities of the pyrimidines in DNA. The complete structural optimizations and harmonic vibrational analyses of dAdT and dAdT- are perhaps the largest reported in no symmetry using DFT methods and good basis sets including both polarization and diffuse functions. Acknowledgment. This research was supported by the National Science Foundation, Grant CHE-0451445, and by the Department of Energy’s National Energy Research Scientific Computing Center (NERSC, Lawrence Berkeley Laboratory). J.G. was supported by the “Knowledge Innovation Program” and the “Introducing Outstanding Overseas Scientists Project” of the Chinese Academy of Sciences. Supporting Information Available: Cartesian coordinates for the optimized neutral dAdT and dAdT- (Tables S1-S2); Harmonic vibrational frequencies and their infrared intensities for dAdT and dAdT- (Tables S3-S4). This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Steenken, S.; Telo, J. P.; Novais, H. M.; Candeias, L. P. J. Am. Chem. Soc. 1992, 114, 4701.

Gu et al. (2) (a) Colson, A. O.; Sevilla, M. D. Int. J. Radiat. Biol. 1995, 67, 627. (b) Li, X.; Sevilla, M. D.: Sanche, L. J. Phys. Chem. B 2004, 108, 5472. (c) Li, X.; Sanche, L.; Sevilla, M. D. J. Phys. Chem. B 2004, 108, 19013. (3) (a) Anusiewicz, I.; Berdys-Kochanska, J.; Sobczyk, M.; Skurski, P.; Simons, J. J. Phys. Chem. A 2004, 108, 11381. (b) Anusiewicz, I.; Sobczyk, M.; Berdys-Kochanska, J.; Skurski, P.; Simons, J. J. Phys. Chem. A 2005, 109, 484. (4) Desfrancois, D.; Abdoul-Carime, H.; Schermann, J. P. J. Chem. Phys. 1996, 104, 7792. (5) Kelley, S. O.; Barton, J. K. Science 1999, 283, 375. (6) Ratner, M. Nature 1999, 397, 480. (7) Hall, D. B.; Holmlin, R. E.; Barton, J. K. Nature 1996, 382, 731. (8) Steenken, S. Biol. Chem. 1997, 378, 129. (9) Taubes, G. Science 1997, 275, 1420. (10) Cai, Z.; Sevilla, M. D. J. Phys. Chem. B 2000, 104, 6942. (11) Messer, A.; Carpenter, K.; Forzley, K.; Buchanan, J.; Yang, S.; Razskazovskii, Y.; Cai, Z.; Sevilla, M. D. J. Phys. Chem. B 2000, 104, 1128. (12) (a) Giese, B.; Spichty, M. Chem. Phys. Chem. 2000, 1, 195. (b) Giese, B.; Spichty, M.; Wessely, S. Pure Appl. Chem. 2001, 73, 449. (c) Giese, B.; Carl, B.; Carl, T.; Carell, T.; Behrens, C. Angew. Chem., Int. Ed. 2004, 43, 1848. (13) Carell, T.; Behrens, C.; Gierlich, J. Org. Biol. Chem. 2003, 1, 2221. (14) Berlin, Y. A.; Burin, A. L.; Ratner, M. A. J. Am. Chem. Soc. 2001, 123, 260. (15) Bixon, M.; Jortner, J. J. Phys. Chem. A 2001, 105, 10322. (16) Huels, M. A.; Hahndorf, I.; Illengerger, E.; Sanche, L. J. Chem. Phys. 1998, 108, 1309. (17) Boudaiffa, B.; Cloutier, P.; Hunting, D.; Huels, M. A.; Sanche, L. Science 2000, 287, 1658. (18) Pan, X.; Cloutier, P.; Hunting, D.; Sanche, L. Phys. ReV. Lett. 2003, 90, 208102. (19) Caron, L. G.; Sanche, L. Phys. ReV. Lett. 2003, 91, 113201. (20) Zheng, Y.; Cloutier, P.; Hunting, D.; Wagner, J. R.; Sanche, L. J. Am. Chem. Soc. 2004, 126, 1002-1003. (21) Barrios, R.; Skurski, P.; Simons, J. J. Phys. Chem. B 2002, 106, 7991. (22) (a) Amann, N.; Pandurski, E.; Fiebig, T.; Wagenknecht, H.-A. Angew. Chem., Int. Ed. 2002, 41, 2978. (b) Amann, N.; Pandurski, E.; Fiebig, T.; Wagenknecht, H.-A. Chem. Eur. J. 2002, 8, 4877. (c) Wagenknecht, H.-A. Angew. Chem., Int. Ed. 2003, 42, 2454. (23) Wiley J. R.; Robinson, J. M.; Chen, E. C. M.; Chen, E. S. D.; Wentworth, W. E. Biochem. Biophys. Res. Commun. 1991, 180, 841. (24) Hendricks, J. H.; Lyzpustina, S. A.; de Clercq, H. L.; Bowen, K. H. J. Chem. Phys. 1996, 104, 7788. (25) Desfrancois, D.; Perquet, V.; Bouteiller, Y.; Schermann, J. P. J. Phys. Chem. A 1998, 102, 1274. (26) Schiedt, J.; Weinkauf, R.; Neumark, D. M.; Schlag, E. W. Chem. Phys. 1998, 239, 511. (27) (a) Chen, E. C. M.; Chen, E. S. D.; Wentworth, W. E. Biochem. Biophys. Res. Commun. 1990, 171, 97. (b) Chen, E. C. M.; Chen, E. S. D. J. Phys. Chem. B 2000, 104, 7835. (28) Periquet, V.; Moreau, A.; Carles, S.; Schermann, J. P.; Desfrancois, C. J. Electron Spectrosc. Relat. Phenom. 2000, 106, 141. (29) Hendricks, J. H.; Lyzpustina, S. A.; de Clercq, H. L.; Bowen, K. H. J. Chem. Phys. 1998, 108, 8. (30) Oyler, N. A. Adamowicz, L. J. Phys. Chem. 1993, 97, 11122. (31) Desfrancois, C.; Abdoul-Carime, H.; Schulz, C. P.; Schermann, J. P. Science 1995, 269, 1707. (32) Nir, E.; Kleinermanns, K.; de Vries, M. S. Nature 2000, 408, 949. (33) Nir, E.; Imhof, P.; Kleinermanns, K.; de Vries, M. S. J. Am. Chem. Soc. 2000, 122, 8091. (34) Nir, E.; Muller, M.; Grace, L. I.; de Vries, M. S. Chem. Phys. Lett. 2002, 355, 59. (35) Nir, E.; Plutzer, C.; Kleinermanns, K.; de Vries, M. S. Eur. Phys. J. D 2002, 20, 317. (36) Rozenberg, M.; Jung, C.; Shoham, G. Phys. Chem. Chem. Phys. 2003, 5, 1533. (37) Freitas, M. A.; Shi, S. D.; Hendrickson, C. L.; Marshall, A. G. J. Am. Chem. Soc. 1998, 120, 10187. (38) Sevilla, M. D.; Besler, B.; Colson, A. J. Phys. Chem. 1995, 99, 1060. (39) Desfrancois, C.; Abdoul-Carime, H.; Carles, S.; Periquet, V.; Schermann, J. P.; Smith, D. M. A.; Adamowicz, L. J. Chem. Phys. 1999, 110, 11876. (40) Smith, D. M. A.; Jalbout, A. F.; Smets, J.; Adamowicz, L. Chem. Phys. 2000, 260, 45. (41) Wetmore, S. D.; Boyd, R. J.; Eriksson, L. A. Chem. Phys. Lett. 2000, 322, 129. (42) Dolgounitcheva, O.; Zakrzewski, V. G.; Ortiz, J. V. J. Phys. Chem. A 2001, 105, 8782.

Feature Article (43) Wesolowski, S. S.; Leininger, M. L.; Pentchev, P. N.; Schaefer, H. F. J. Am. Chem. Soc. 2001, 123, 4023. (44) Li, X.; Sevilla, M. D.; Sanche, L. J. Am. Chem. Soc. 2003, 125, 8916. (45) Russo, N.; Roscano, M.; Grand, A. J. Comput. Chem. 2000, 21, 1243. (46) Rienstra-Kiracofe, J. C.; Tschumper, G. S.; Schaefer, H. F.; Nandi, S.; Ellison, G. B. Chem. ReV. 2002, 102, 231. (47) Colson, A. O.; Besler, B.; Sevilla, M. D. J. Phys. Chem. 1992, 96, 9787. (48) Al-Jihad, I.; Smets, J.; Adamowicz, L. J. Phys. Chem. A 2000, 104, 2994. (49) Smets, J.; Jalbout, A. F.; Adamowicz, L. Chem. Phys. Lett. 2001, 342, 342. (50) Saettel, N. J.; Wiest, O. J. Am. Chem. Soc. 2001, 123, 2693. (51) Smets, J.; Smith, D. M. A.; Elkadi, Y.; Adamowicz, L. J. Phys. Chem. A 1997, 101, 9152. (52) Dolgounitcheva, O.; Zakrzewski, V. G.; Ortiz, J. V. J. Phys. Chem. A 1999, 103, 7912. (53) Li, X.; Cai, Z.; Sevilla, M. D. J. Phys. Chem. B 2001, 105, 10115. (54) Li, X.; Cai, Z.; Sevilla, M. D. J. Phys. Chem. A 2002, 106, 9345. (55) Reynisson, J.; Steenken, S. Phys. Chem. Chem. Phys. 2002, 4, 5353. (56) Kumar, A.; Knapp-Mohammandy, M.; Mishra, P. C.; Suhai, S. J. Comput. Chem. 2004, 25, 1047. (57) (a) Richardson, N. A.; Wesolowski, S. S.; Schaefer, H. F. J. Am. Chem. Soc. 2002, 124, 10163. (b) Richardson, N. A.; Wesolowski, S. S.; Schaefer, H. F. J. Phys. Chem. B 2003, 107, 848. (58) Harris, D. G.; Shao, J.; Anderson, J. M.; Marx, D. P.; Zimmerman, S. S. Nucleosides, Nucleotides, Nucleic Acids 2002, 21, 803. (59) Voityuk, A. A.; Michel-Beyerle, M.; Rosch, N. Chem. Phys. Lett. 2001, 342, 231. (60) Richardson, N. A.; Gu, J.; Wang, S.; Xie, Y.; Schaefer, H. F. J. Am. Chem. Soc. 2004, 126, 4404. (61) Steenken, S. Chem. ReV. 1989, 89, 503. (62) Bertran, J.; Oliva, A.; Rodriquez-Santiago, L.; Sodupe, M. J. Am. Chem. Soc. 1998, 120, 8159. (63) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (64) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785. (65) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.

J. Phys. Chem. B, Vol. 109, No. 27, 2005 13075 D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision a.10; Gaussian, Inc.: Pittsburgh, PA, 1998. (66) Huzinaga, S. J. Chem. Phys. 1965, 42, 1293. (67) Dunning, T. H. J. Chem. Phys. 1970, 53, 2823. (68) Lee, T. J.; Schaefer, H. F. J. Chem. Phys. 1985, 83, 1784. (69) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985, 83, 735. (70) Reed, A. E.; Weinhold, F. J. Chem. Phys. 1985, 83, 1736. (71) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. 1988, 88, 899. (72) Reed, A. E.; Schleyer, P. v. R. J. Am. Chem. Soc. 1990, 112, 1434. (73) Huels, M. A.; Boudaiffa, B.; Cloutier, P.; Hunting, D.; Sanche, L. J. Am. Chem. Soc. 2003, 125, 4467. (74) Zheng, Y.; Cloutier, P.; Hunting, D. J.; Wagner, J. R.; Sanche, L. J. Am. Chem. Soc. 2004, 126, 1002. (75) Gu, J.; Xie, Y.; Schaefer, H. F. J. Am. Chem. Soc. 2005, 127, 2818. (76) Abdoul-Carime, H.; Gohlke, S.; Fischbach, E.; Scheike, J.; Illenberger, E. Chem. Phys. Lett. 2004, 387, 267. (77) Barrios, R.; Skurski, P.; Simons, J. J. Phys. Chem. B 2002, 106, 7991. (78) Berdys, J.; Ansiewicz, I.; Skurski, P.; Simons, J. J. Phys. Chem. A 2004, 108, 2999. (79) Berdys, J.; Skurski, P.; Simons, J. J. Phys. Chem. B 2004, 108, 5800. (80) Berdys, J.; Ansiewicz, I.; Skurski, P.; Simons, J. J. Am. Chem. Soc. 2004, 126, 6441. (81) Li, X.; Sevilla, M. D.; Sanche, L. J. Am. Chem. Soc. 2003, 125, 13668. (82) Smith, B. J.; Swanton, D. J.; Pople, J. A.; Schaefer, H. F. J. Chem. Phys. 1990, 92, 1240. (83) Sponer, J.; Jurecka, P.; Hobza, P. J. Am. Chem. Soc. 2004, 126, 10142. (84) Spirko, V.; Sponer, J.; Hobza, P. J. Chem. Phys. 1997, 106, 1472.