6574
J. Phys. Chem. B 2009, 113, 6574–6578
Hydrogen Abstraction from Deoxyribose by a Neighboring 3′-Uracil Peroxyl Radical Patric Schyman,*,† Leif A. Eriksson,*,‡ and Aatto Laaksonen† DiVision of Physical Chemistry, Arrhenius Laboratory, Stockholm UniVersity, 106 91 Stockholm, Sweden, ¨ rebro Life Science Center, School of Science and Technology, O ¨ rebro UniVersity, and O ¨ Fakultetsgatan 1, 701 82 Orebro, Sweden ReceiVed: January 26, 2009; ReVised Manuscript ReceiVed: March 5, 2009
Theoretical examination of the reactivity of the uracil-5-peroxyl radical when abstracting a hydrogen atom from a neighboring 5′-deoxyribose in 5′-ApU-5-peroxyl-3′ has been performed using density functional theory with the MPWB1K functional. Halogenated uracils are often used as radiosensitizers in DNA since the reactive uracil-5-yl radical is formed upon radiation and is known to create strand break and alkali-labile sites. Under aerobic conditions, such as in the cell, it has been proposed that the uracil-5-peroxyl radical is formed and would be the damaging agent. Our results show low reactivity for the uracil-5-peroxyl radical, determined by calculating the activation and reaction energies for the plausible hydrogen abstraction sites C1′, C2′, and C3′ of the neighboring 5′-deoxyribose. These findings support the hypothesis that hydrogen abstraction primarily occurs by the uracil-5-yl radical, also under aerobic conditions, prior to formation of the peroxyl radical. Introduction
SCHEME 1
Radiosensitizers, as used in radiotherapy, make tumor cells more sensitive to radiation exposure whereby less radiation is needed to damage the malignant cells, thus also affecting the neighboring healthy tissue to a less extent.1,2 The halogenated pyrimidines 5-bromouracil (BrU) and 5-fluorouracil (FU) are efficient radiosensitizers that can substitute thymine in DNA and thereby make DNA more vulnerable toward UVB and ionizing radiation. The substitution is enabled by the comparable sizes of the halogen atoms and the methyl group in thymine. DNA incorporating halogenated pyrimidines is more vulnerable because they are easily dehalogenated by radiation thereby creating reactive uracil-5-yl radicals3-5 that can induce strand break and alkali-labile sites in DNA.3 The DNA damage prevents accurate DNA replication and can eventually lead to apoptosis of the malignant cells.2 The dehalogenation process for BrU can occur in several ways. Upon exposure to UVB radiation, the uracil-5-yl radical can be formed from either homolytic cleavage of the Br-U bond during photolysis or intramolecular electron transfer from an adjacent base. In this context, a sequence selectivity has been noted, in that neighboring adenine bases are the preferred sites of oxidation.3,6-8 The homolytic bond cleavage generates a reactive Br · radical and the uracil-5-yl radical, while electron transfer generates Br-, the uracil-5-yl radical, and an adjacent base radical cation. Br · is in turn often responsible for oxidation of neighboring bases.3 When exposed to ionizing radiation, BrU is reduced to BrU · - by addition of solvated electrons generated by radiolysis, or by electron transfer from neighboring base radical anions.3 In these cases the result is again a uracil-5-yl radical and a Br- ion. The generated uracil-5-yl radical will most likely abstract a hydrogen atom from a neighboring 5′-deoxyribose. Experimental8-12 and theoretical13 studies of B-DNA under anaerobic
conditions (absence of oxygen) indicate that hydrogen abstraction can occur at either the C1′ or the C2′ site on the neighboring deoxyribose, with a preference for the C1′ site. However, in the presence of oxygen, which is the condition in the cell, the uracil-5-peroxyl radical can be formed;3,14 cf. Scheme 1. Doddridge et al. conducted UVB experiments on single-stranded DNA (ssDNA) under aerobic conditions.14 They argued that the uracil-5-peroxyl radical predominantly abstracts the hydrogen atom from C1′. Very few studies have, however, been devoted to the hydrogen abstraction ability of the uracil-5-peroxyl radical compared to the parent uracil-5-yl radical. In fact, it has been suggested that even in the presence of oxygen the uracil-5-yl radical is so reactive that it will abstract a hydrogen atom from the neighboring deoxyribose before the uracil-5-peroxyl radical has been formed.15 We thus decided to examine the possibility of hydrogen abstraction by the uracil-5-peroxyl radical computationally, using a 5′-ApU-5-peroxyl-3′ model, cf. Scheme 2, in order to evaluate previous experimental observations.14 Possible hydrogen abstraction sites on the deoxyribose of the neighboring nucleoside are C1′, C2′, and C3′; see Scheme 2.
* To whom correspondence should be addressed. Fax: +46 19 303566; e-mail:
[email protected] (L.A.E.). E-mail:
[email protected] (P.S.). † Stockholm University. ‡ ¨ Orebro University.
Method All geometries were optimized using the MPWB1K functional16-18 as implemented in Gaussian 03.19 Geometry optimizations were performed with the double-ζ basis set 6-31G(d,p) in vacuum. During the optimizations no atoms were held fixed or restricted to move. The transition-state searches were conducted using the QST320 algorithm as implemented in Gaussian 03 and verified by performing frequency calculations using the same level of theory as used for optimization. Unless explicitly mentioned, the electronic energies were obtained from single-
10.1021/jp9007569 CCC: $40.75 2009 American Chemical Society Published on Web 04/09/2009
Reactivity of the Uracil-5-peroxyl Radical
J. Phys. Chem. B, Vol. 113, No. 18, 2009 6575
SCHEME 2
point calculations using the basis set 6-311+G(2d,2p). To take into account the surroundings, a dielectric model (IEF-PCM)21 was used in single-point calculations with a dielectric constant of ε ) 78.4 to model bulk water, and also with a lower dielectric constant (ε ) 4) to simulate a more nonpolar environment. In the IEF-PCM calculations, the larger basis set was used. The DNA model (5′-ApT-3′) was obtained from a crystal structure of B-DNA (the coordinates were taken from the Protein Data Bank, code 1bdna) in which the thymine base was modified by replacing its methyl group with O2, thus creating a uracil5-peroxyl radical, 5′-ApU-5-peroxyl-3′; cf. Scheme 2. The phosphate group was neutralized by protonating one of the oxygen atoms to simulate a closely located counterion. In a recent study, it was shown that the protonation state of the phosphate group does not significantly influence the relative energetic differences when abstracting the C1′ or C2′ hydrogens.22 Results The B3LYP functional23,24 is known to have problems with accurately describing π interactions and dispersion forces such as those occurring between nucleobases in DNA.25 The more recent MPWB1K functional,16-18 developed by Truhlar and coworkers, aims to overcome some of the shortcomings of B3LYP.26 In an earlier paper,13 where we studied the uracil-5yl radical in a similar model (5′-ApU-5-yl-3′), the B3LYPoptimized geometry was distorted and did not show the typical π-stacked geometry, as can be seen in Figure 1. This intrigued us to reoptimize our old model using the MPWB1K functional. The geometries of both optimizations are shown in Figure 1.
The geometry obtained with the new functional shows closely packed nucleobases typical for π interactions. Figure 1 also shows, to the right, the optimized geometry of the uracil-5peroxyl radical and that the π interactions remain. From here on only the MPWB1K geometries will be discussed unless otherwise explicitly mentioned. In Figure 2 the distances between the peroxyl radical center and possible hydrogen abstraction sites are marked. The differences in distances between the models 5′-ApU-yl-3′ and 5′-ApU-peroxyl-3′ are less than 0.6 Å. In both cases the C2′ hydrogen is the closest with a distance of ∼3 Å while C1′ hydrogen is further away from the radical center of the peroxyl in the 5′-ApU-peroxyl-3′ model. However, the peroxyl radical is more mobile than the C5 centered uracilyl radical, since the C5-O(O) rotation energy is low (cf. Figure 3). The highest energy is in fact reached when the oxygen is rotating away from the C1′, C2′, and C3′ hydrogens, 360°-310°. Comparison of the Mulliken spin densities at the radical centers indicates that the uracil-5-peroxyl radical is less reactive than the uracil-5-yl radical. The uracil-5-peroxyl radical unpaired spin density is delocalized over the peroxyl oxygen atoms with 0.74 e- at the terminal oxygen atom (Oouter) and 0.22 e- at the inner one. In aqueous solution, the values are essentially unaltered. The values are considerably lower than the spin density of 1.3 e- found at the radical center C5 of the uracil5-yl radical. C1′-H Abstraction. The distance between the C1′ hydrogen and the peroxyl radical is 4.1 Å. The TS guess was obtained by performing an energy scan in which the distance between the C1′ hydrogen and the terminal peroxyl oxygen atom was
6576
J. Phys. Chem. B, Vol. 113, No. 18, 2009
Schyman et al.
Figure 1. Optimized geometries showing the difference in base stacking obtained with the B3LYP and MPWB1K functionals.
Figure 2. Stationary structures for the three hydrogen abstraction paths.
gradually reduced. The geometry corresponding to the highest energy on the energy profile was then used in the QST3 algorithm to search for a saddle point. In the optimized TS the hydrogen atom is located 1.23 Å from Oouter and 1.33 Å from C1′; see Figure 2. The frequency calculation verified the TS, with an imaginary frequency of 2213.1i cm-1.
The activation energy for the hydrogen atom transfer was calculated by taking the energy difference between the TS and the reactant, and correcting for the zero-point energy; in Table 1 all energies are compiled. In the gas phase the activation energy for the C1′ hydrogen atom transfer is 24.5 kcal/mol. When a solvation model with a dielectric constant of water is
Reactivity of the Uracil-5-peroxyl Radical
J. Phys. Chem. B, Vol. 113, No. 18, 2009 6577
Figure 3. Energy scan of rotating the OOuter oxygen about the OInner-C5U bond.
TABLE 1: Energies Relative to the Reactant (kcal mol-1)a 6-31G(d) gas phase TS-C1′ TS-C2′ TS-C3′ product C1′* product C2′* product C3′*
26.72 31.36 25.07 10.12 18.51 16.83
b
6-311+G(2d,2p) gas phase
ε)4
ε ) 78
24.51 28.35 21.69 6.78 14.49 12.74
21.82 27.09 20.43 6.16 12.46 12.66
20.78 26.18 20.47 6.36 11.11 13.19
a All TS energies include zero-point energy corrections. nic energy.
b
Electro-
applied, the activation energy drops to 20.8 kcal/mol. In order to achieve this hydrogen atom transfer, a major structural change is necessary as can be seen in Figure 2. The 5′-deoxyribose is rotated such that the plane of the sugar ring is orthogonal to the plane of the 3′-deoxyribose. These planes are normally close to parallel. The C1′ · deoxyribose radical formed after the C1′ hydrogen abstraction still has the rotated 5′-deoxyribose which in part can be explained by the hydrogen bond formed between the terminal peroxylic hydrogen and one of the oxygens in the phosphate group, with a hydrogen bond distance of 1.79 Å. The product is 6.8 kcal/mol less stable than the reactant in the gas phase. C2′-H Abstraction. The search for the C2′ hydrogen atom transfer TS was conducted in a similar manner as for the C1′ transition state. The TS found is very late, with the hydrogen atom almost attaining a normal H-O bond length to the peroxyl oxygen (Figure 2). At the TS the distances are C2′-H ) 1.48 Å and H-Oouter ) 1.10 Å, and the imaginary frequency 1757.7i cm-1. When going from the reactant to the TS, only small geometry changes are observed, in comparison with the C1′ hydrogen atom transfer. The reason for the moderate geometry changes is that the peroxyl radical moiety in its rotational minimum configuration is well positioned for attack in the optimized 5′-ApU-peroxyl-3′ model, with the C2′ hydrogen located only 2.9 Å from Oouter. The energy barrier relative to the reactant for abstracting the C2′ hydrogen by the uracil-5peroxyl was in this case 28.4 kcal/mol in the gas phase; see Table 1. In aqueous environment the energy is slightly lower, 26.2 kcal/mol. The molecule does not undergo any major structural changes when forming the C2′ radical. A hydrogen bond is also formed in this case between the terminal peroxylic hydrogen and one of the oxygens in the phosphate group. The energy relative to the reactant is 14.5 kcal/mol in the gas phase, which is 7.7 kcal/
mol higher than the C1′ radical product. Both the higher barrier and the more endothermic reaction energy indicate that the C2′ radical will be formed in less yield than the C1′ radical. C3′-H Abstraction. The C3′-H is the hydrogen furthest away from the uracil-5-peroxyl radical oxygen of the three considered here, at a distance of 5.2 Å. The C3′ hydrogen abstraction was not considered in our earlier study with the uracil-5-yl radical since that would have required major structural changes. The peroxyl radical considered in the current study becomes an extension of the radical center enabling the C3′ hydrogen to come within reach. The TS is, just as for the C2′ system, asymmetric with the hydrogen atom close to the product state (see Figure 2) and has an imaginary frequency of 2042.9i cm-1. The distances for the hydrogen atom transfer TS are C3′-H ) 1.39 Å and H-Oouter ) 1.16 Å. The energy barrier for the C3′ hydrogen atom transfer is 21.7 kcal/mol in the gas phase and 20.5 kcal/mol in water. To reach the TS the system needs to undergo a structural change as can be seen in Figure 2, although the geometric distortion in this case is less than that seen for the C1′-H abstraction. The most significant geometry change needed in order to position the C3′-H for abstraction is a rotation about the (P)O-C3′ bond, in turn causing the 5′-deoxyribose to rotate away from the uracil5-peroxyl moiety. In the rotation, the π-stacking between the bases is strongly affected. After the hydrogen atom transfer the deoxyriboses of the C3′ radical product are still distorted. Again, it is possible that the hydrogen bond formed between the phosphate oxygen and the peroxyl hydrogen (distance 1.83 Å) hinders the geometry to return to its original shape; see Figure 2. The energy of the C3′ radical product relative to the reactant is 12.7 kcal/mol in the gas phase. Interestingly, both the C1′ and C3′ products have lower energy than the C2′ product in the gas phase, despite the fact that they involve considerably larger structural changes. In aqueous solution, however, the reaction energy of the C2′ product lies between the two others. Discussion The reactant geometry of the 5′-ApU-5-peroxyl-3′ radical and the rotational energy surface of the peroxyl group show that the terminal peroxyl oxygen is readily positioned for hydrogen abstraction from all three sites C1′, C2′, and C3′ of the neighboring 5′-deoxyribose. From visual inspection the C1′ and C3′ hydrogen abstractions may appear more probable for the uracil5-peroxyl radical than for the uracil-5-yl radical since the whole nucleobase would in the latter case have to rotate in order to
6578
J. Phys. Chem. B, Vol. 113, No. 18, 2009
reach the C1′ or C3′ hydrogens; in the peroxyl radical the extension of the radical center due to the two oxygens will geometrically facilitate the hydrogen atom transfer. However, as shown in Figure 2, major structural changes are necessary to reach the TS’s, more so for the C1′-H abstraction than the C3′-H abstraction, also in the peroxyl radical cases. This kind of deformation would most likely not be feasible in doublestranded DNA (dsDNA). It may still be possible in ssDNA, as observed experimentally by Doddridge et al.14 The main reason for the geometry change originates from the mobility of the 5′-deoxyribose and the adenine base. It cannot be ruled out at this point that some of the geometrical changes could be hindered when extended to a full DNA sequence. The only hydrogen abstraction that did not need any major geometry change was the C2′-H abstraction, due to the relatively close proximity between the reacting sites already from the onset (the C2′ hydrogen is in the reactant located only 3 Å from the peroxyl radical). From a structural analysis, the C2′ site is hence the most attractive abstraction site in a compact geometry as would be the case in dsDNA. Analysis of the energies in Table 1 shows that the C2′ hydrogen abstraction has a very high energy barrier and is in fact the least probable abstraction site of those considered herein. More importantly, the energies reveal very low reactivity for the uracil-5-peroxyl radical compared with the uracil-5-yl radical species.13 For the uracil-5-yl radical C1′ hydrogen abstraction the energy barrier in water was even negative, -2.5 kcal/mol (energies obtained at the BB1K/6-31+G(d,p) level),13 whereas for the uracil-5-peroxyl radical the corresponding energy barrier is 20.8 kcal/mol. Albeit obtained using a different method and basis set, this effect should be a few kcal/mol at most. The corresponding energy difference for the C2′ hydrogen abstraction is almost the same: 2.8 kcal/mol for the uracil-5-yl radical and 26.2 kcal/mol for the uracil-5-peroxyl radical. This provides some insight into the high efficiency of the uracil-5-yl radical in abstracting a hydrogen atom from the neighboring deoxyribose compared to other hydrogen abstraction radicals.13,22 It was here shown that the C3′ hydrogen abstraction has the lowest barrier; it would therefore be of interest to explore the energy of C3′-H abstraction by the uracil-5-yl radical in a future study. The energies of the uracil-5-peroxyl products, the deoxyribose C1′, C2′, and C3′ radicals, are rather high, 6, 11, and 13 kcal/ mol, respectively. This should be compared with the energies of the uracil-5-yl products, the deoxyribose C1′ and C2′ radicals, found at -30.3 and -20.6 kcal/mol, respectively, in aqueous solution. Hence, the uracil-5-yl radical does not only react more rapidly, but the products formed are also far more stable. Conclusions Our calculations show that, for the 5′-ApU-5-peroxyl-3′ model, the lowest energy barrier for hydrogen abstraction is found for the C3′ site, although the C1′ site is almost as low in energy, while the C2′ site has the highest energy barrier of all three. The energy difference between C1′ and C3′ hydrogen abstraction decreases in aqueous solution, suggesting that the smallest energy difference should be seen in ssDNA. The proximity of the terminal peroxyl oxygen to the C1′ hydrogen could explain the experimentally observed C1′ hydrogen abstraction in ssDNA.14 Judging from the structural changes required at the TS’s for the C1′ and C3′ H-abstractions, it is unlikely that these reactions would take place in dsDNA. It is possible that the approximately 6 kcal/mol higher C2′ TS in
Schyman et al. water will become more favorable in a rigid dsDNA structure. Overall, both barrier heights and reaction energies point to that the peroxyl radical will be less efficient in abstracting H atoms from the deoxyribose units than the parent uracil-5-yl radical. Acknowledgment. The Swedish Science Foundation (VR) ¨ rebro University and the Faculty of Science and Technology at O are gratefully acknowledged for financial support. We also acknowledge grants of computing time at the National Supercomputing Center (NSC) in Linko¨ping and the Center for Parallel Computing (PDC) at the Royal Institute of Technology. Supporting Information Available: The optimized geometries of all species discussed herein. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Rutman, R. J.; Cantarow, A.; Pashkis, K. E. Cancer Res. 1954, 14, 119. (2) Longley, D. B.; Harkin, D. P.; Johnston, G. P. Nature (London) 2003, 3, 330. (3) von Sonntag, C. Free-Radical-Induced DNA Damage and Its Repair; Schreck, S., Ed.; Springer-Verlag: Berlin, 2006. (4) Li, X.; Sanche, L.; Sevilla, M. D. J. Phys. Chem. A 2002, 106, 11248. (5) Wetmore, S. D.; Boyd, R. D.; Eriksson, L. A. Chem. Phys. Lett. 2001, 343, 151. (6) Chen, T.; Cook, G. P.; Koppisch, A. T.; Greenberg, M. M. J. Am. Chem. Soc. 2000, 122, 3861. (7) Sugiyama, H.; Tsutsumi, Y.; Saito, I. J. Am. Chem. Soc. 1990, 112, 6720. (8) Saito, I. Pure Appl. Chem. 1992, 64, 1305. (9) Watanabe, T.; Bando, T.; Xu, Y.; Tashiro, R.; Sugiyama, H. J. Am. Chem. Soc. 2004, 127, 44. (10) Cook, G. P.; Greenberg, M. M. J. Am. Chem. Soc. 1996, 118, 10025. (11) Sugiyama, H.; Fujimoto, K.; Sito, I. Tetrahedron Lett. 1996, 37, 1805. (12) Fujimoto, K.; Ikeda, Y.; Ishihara, S.; Saito, I. Terahedron Lett. 2003, 43, 2243. (13) Schyman, P.; Zhang, R. B.; Eriksson, L. A.; Laaksonen, A. Phys. Chem. Chem. Phys. 2007, 9, 5975. (14) Doddridge, Z.; Warner, J. L.; Cullis, P. M.; Jones, G. D. D. Chem. Commun. 1998, 1997. (15) von Sonntag C. The Chemical Basis of Radiation Biology; Taylor & Francis: London, 1987. (16) Zhao, Y.; Truhlar, D. G. J. Phys. Chem. A 2004, 108, 6908. (17) Becke, A. D. J. Chem. Phys. 1996, 98, 1040. (18) Adamo, C.; Barone, V. J. Phys. Chem. 1998, 108, 664. (19) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, ReVision B.04; Gaussian Inc.: Pittsburgh, PA, 2003. (20) Ayala, P. Y.; Schlegel, H. B. J. Chem. Phys. 1997, 107, 375. (21) Tomasi, J.; Mennucci, B.; Cammi, R. Chem. ReV. 2005, 105, 2999. (22) Schyman, P.; Eriksson, L. A.; Zhang, R. B.; Laakonen, A. Chem. Phys. Lett. 2008, 458, 186. (23) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785. (24) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (25) Hobza, P.; Sponer, J. Chem. ReV. 1999, 99, 3247. (26) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157.
JP9007569
