Structural Origin of Copper Ion Containing Artificial DNA: A Density

Dec 2, 2008 - a lack of dispersion forces when one adopts the ordinary DFT and because ... Boys-Bernardi's counterpoise method.34 The interaction ener...
2 downloads 0 Views 249KB Size
16960

J. Phys. Chem. B 2008, 112, 16960–16965

Structural Origin of Copper Ion Containing Artificial DNA: A Density Functional Study Toru Matsui,† Hideaki Miyachi,† Takeshi Sato,† Yasuteru Shigeta,‡ and Kimihiko Hirao*,† Department of Applied Chemistry, School of Engineering, The UniVersity of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan, and Graduate School of Life Science, UniVersity of Hyogo, 3-2-1 Kouto, Kamigouri-cho, Ako-gun, Hyogo 678-1297, Japan ReceiVed: September 11, 2008; ReVised Manuscript ReceiVed: October 15, 2008

In order to investigate an origin of structural stability of a copper ion containing artificial DNA, we evaluated the stacking energy between [H-Cu2+-H] (H: hydroxypyridone) dimer by means of density functional theory (DFT) with an Anderson-Langreth-Lindqvist van der Waals (vdW) functional. The calculated distance between the copper ions is about 3.6 Å, which agrees well with the experimental data. Evaluated stacking energy is about 8-10 kcal/mol, which is slightly smaller than that of two base pairs in a natural B-DNA. This tendency does not change in [H-2H+-H], which does not contain copper ions. These results indicate that the vdW interaction dominates the inter-base-pair interaction over spin-spin interaction, in contrast to a conjecture by an experimental group. According to the results by the open-shell DFT, antiferromagnetic (singlet) and ferromagnetic (triplet) states are almost degenerated when the two bases are vertically located and both bases have a planar structure as found in the B-DNA. 1. Introduction Many studies all over the world have been trying to apply the DNA as a functional material toward nanotechnology, since the DNA has several stiff architectures such as a duplex (BDNA form) and a guanine quadruplex. These architectures are suitable for an alignment of bases along a direction. Some studies reported that a guanine triplet (a GGG sequence) tends to undergo a one-electron oxidation reaction, and the electron can move through base pairs. This property and mechanism have been explained both experimentally1-3 and theoretically4-6 in relation to a DNA damage and repair. Other studies have tried to observe electronic conductivity through π-π stacking between base pairs.7-10 The conventional calculation11 reported that the conductivity is large when the electrodes are connected to a base moiety of the DNA. On the other hand, it also reported that the conductivity goes down dramatically when the electrodes are connected to the backbone of the DNA. Therefore, the question of whether the DNA has a potential for electron conductive material or not remains unsolved. In such a circumstance, it is an interesting idea to make some metal complexes bind to a DNA for the purpose of controlling the structural and electronic properties of the DNA. There are several kinds of systems which contain a metal complex and the base moiety of the DNA. For instance, a cisplatin, which is known as an antitumor drug, distorts the structure of the B-DNA by 90°.12,13 In this case, Pt2+ of the cisplatin prefers to bind to N7 of a guanine to that of an adenine. Like the platinum complexes, many metal complexes containing metal ions such as Ni2+, Zn2+, and Mg2+ bind to N7 of the guanine.14-17 In other cases like a metal-complexed DNA (M-DNA),18,19 a zinc complex binds between a guanine and a cytosine (an adenine and a thymine), where the metal complex has a bridge structure. Moreover, in 2006, Miyake et al. succeeded to make Hg2+ binded to a thymine-thymine (T-T) mismatch which resulted * Corresponding author: Fax: 81 3 5841 7241; e-mail: hirao@ qcl.t.u-tokyo.ac.jp. † The University of Tokyo. ‡ University of Hyogo.

in a [T-Hg2+-T] complex.20 However, it is difficult to make a programmable way to bind only the metal together with the DNA by considering metal-DNA interaction. The computational chemistry is good at observing changes in the base moieties of the DNA caused by a metal complex binding. In our previous study, we revealed that a platinum complex promotes the charge separation in a guanine-cytosine pair.21,22 Other papers also reported that the same phenomenon is observed in the cases of the other metal complexes.23 Voityuk showed that a [T-Hg2+-T] pair plays an important role in an excess electron transfer in DNA by comparing electron-transfer coupling constants before and after the metal binding, where the electron hopping rate of [T-Hg2+-T] complexes becomes higher than the natural B-DNA.24 Recently, an artificial DNA, which consists of a DNA-like base capturing the metal ions, sugar, and phosphoric acids, has received much attention.25 The artificial DNA has a selectivity of metal ions, which are captured by DNA-like bases, so that the metal ions could be arrayed hierarchically. In fact, many artificial DNAs with various metal ions have been synthesized and reported.26-28 In 2003, Tanaka succeeded to array five [H-Cu2+-H] (H: hydroxypyridone) into a DNA duplex.29,30 Although details of the structure by NMR or X-ray experiments have not yet been available, they revealed that the distance between the copper ions are 3.7 ( 0.1 Å with electron paramagnetic resonance (EPR) experiments at 1.5 K. Since the available structural information by the experiments is limited, the computational chemistry might contribute to get a deep understanding of the structure of the artificial DNA. In the artificial DNA, it is obvious that metal-ligand interaction (in the case of [H-Cu2+-H], the interaction between Cu2+ and O-) takes over the hydrogen-bonding interaction between Hs as an intra-base-pair interaction. However, what kind of interaction works as the inter-base-pair interaction to form a stable structure has not been clear. In spite of the repulsive Coulomb interaction among the copper ions and ligands, the copper containing artificial DNA is as stable as the natural B-DNA. Some experimental groups suggested that the

10.1021/jp8080707 CCC: $40.75  2008 American Chemical Society Published on Web 12/02/2008

Copper Ion Containing Artificial DNA SCHEME 1: Chemical Structural Formula of cis-[H-Cu2+-H] (a) and trans-[H-Cu2+-H] (b), Where dP Represents the Backbone Molecule

spin-spin interactions among the metal ions should play an important role in the inter-base-pair interaction. Since the order of spin-spin interaction is usually too small in comparison with a chemical bond, it is natural to assume that the inter-base-pair interaction in the artificial DNA is the same as that in the natural B-DNA. Nevertheless, this possibility has not been explored for the artificial DNA yet. Some groups have reported theoretical works on the artificial DNA. Zhang and co-workers investigated the electronic properties of an infinite periodic wire of the artificial DNA bases by a spin-polarized density functional theory (DFT).31 They found that ferromagnetic and antiferromagnetic states are energetically degenerated. Jishi calculated the total energy of [H-Cu2+-H] with three different symmetries.32 They computed the distance between copper ions with the Morse potential. Nevertheless, it has not been clear so far in computational chemistry articles which [H-Cu2+-H] isomers are favorable. Therefore, it is desirable to investigate all possible isomers and choice of models in detail to have a deep knowledge about the structure of the artificial DNA. In this study, in order to elucidate the structural stability of the artificial DNA, we evaluated an interbase distance and a corresponding stacking energy by using the DFT with an Anderson-Langreth-Lindqvist (ALL) vdW functional, which gives reasonable stacking energies of both the benzene dimer and DNA bases as reported previously,33 and examined what is the main contribution to the inter-base-pair interaction in the artificial DNA. This study aims to derive the stacking energy of [H-Cu2+-H] dimer and the understanding of the stacking interaction difference between [H-Cu2+-H] (with copper ions) and [H-2H+-H] (without copper ions). We also investigate a singlet-triplet energy gap. In section 2, we explained computational details and model systems in this work. The numerical results of the stacking energy are given in section 3. The singlet-triplet energy gap is discussed in section 4. Finally the concluding remarks are given in section 5. 2. Computational Details We considered two isomers of [H-Cu2+-H]. One is a cis type whose nitrogen atoms are located on the same side (see Scheme 1a), and the other is a trans type whose nitrogen atoms are located on the opposite side (Scheme 1b). In the actual calculation, we replace the backbone molecules (deoxyribose5′-phosphate (dP)) by hydrogen atoms or methyl groups for simplicity, and then we optimized the geometry of one artificial base pair cis-[H-Cu2+-H] and trans-[H-Cu2+-H]. We found that both the cis-[H-Cu2+-H] and the trans-[H-Cu2+-H] have a planar structure like the base pairs in the natural B-DNA (see Supporting Information). We next evaluated the stacking energy between the base pairs. It is difficult to optimize geometries of two base pairs because no stable structure of two base pairs is found due to a lack of dispersion forces when one adopts the ordinary DFT and because it costs too much to compute gradients by means of the DFT with the vdW correction. Therefore, in order to

J. Phys. Chem. B, Vol. 112, No. 51, 2008 16961 get the potential energy surface, we performed the singlepoint calculations by using the DFT with the vdW correction by fixing the distance between copper ions. In the calculation of two base pairs of [H-Cu2+-H], the upper base is vertically located and twisted by 36° like the bases in the natural DNA base pair (see Figure 1). In order to estimate the interaction energy, we used the Boys-Bernardi’s counterpoise method.34 The interaction energy (Eint) is obtained by the following formula:

Eint ) E(A-B) - {E(a-B) + E(A-b)}

(1)

where E(A-b) means the energy of molecule A obtained with basis sets at the position of both molecule A and molecule B. In order to consider the stacking interaction between base pairs, we used the ALL functional35 for the vdW interaction. Adding the vdW correction, (1) is rewritten as total Eint ) Eint + EvdW

(2)

where EvdW is the energy correction obtained by the ALL functional. The stacking energy means the negative interaction energy, Estacking ) -Eint. We should again stress here that this method well reproduces results for the stacking energy of the B-DNA obtained by high-level ab initio calculations with the cost of the DFT as described in the Introduction.36,37 We adopted the long-range corrected (LC) density functional theory, where the exchange functional is Becke 88 functional and the correlation functional is one-parameter (OP) functional developed in our laboratory. We will refer this method as “LC-BOP”.38,39 Basis sets used here are 6-311+G(d) for the copper atom and 6-31++G(d,p) for the other atoms. Throughout this study, we used a modified version of the GAUSSIAN03 program package.40 3. Results and Discussion 3.1. Choice of Models. In order to confirm the appropriateness of the models for the artificial DNA, we first confirm the model dependence of the interbase pair interaction. We adopt two models: (a) a hydrogen atom model (dP ) H) and (b) a methyl group model (dP ) CH3). These models are often used in the analyses of the inter-base-pair interaction between the bases in the natural B-DNA, where it is known that the backbone molecules do not contribute much to the stability of the duplex structure. We here also make the same assumption for the artificial DNA. As mentioned in the Introduction, we here also focus on the stability of the cis and trans isomers. 3.1.1. Hydrogen Atom and Methyl Group Models. Figure 2 shows the interaction energy between the hydrogen atom model and the methyl group model. Without the vdW correction, the interaction energy decreases as r becomes larger. The interaction seems to be repulsive over whole region mainly due to the repulsive Coulomb interaction between the bases, and no local minimum is found. With the vdW correction, on the other hand, the interaction energy becomes negative and [H-Cu2+-H] planes attract each other. Moreover, the local minimum of the interaction energy was found around to be 3.55-3.65 Å. Thus, it is necessary to include the effect of the vdW interaction to form the DNA-like structure. Next, we discuss the difference of the stacking energy of (a) hydrogen atom model (dP ) H) and (b) the methyl group model (dP ) CH3). The stacking energy of the methyl group model is 1.5-2 kcal/mol larger than that of the hydrogen atom model.

16962 J. Phys. Chem. B, Vol. 112, No. 51, 2008

Matsui et al.

Figure 1. (a) cis-[H-Cu2+-H] dimer used for calculation. (b) Model of (a), where M represents metal cation and r means the distance between copper ions.

Figure 2. Interaction energy between (a) hydrogen atom model (dP ) H) and (b) methyl group model (dP ) CH3). In the figure, Int E of LC-BOP total is obtained by Eint, and LC-BOP+ALL is obtained by Eint .

Figure 3. Interaction energy between (a) cis-[H-Cu2+-H] and (b) trans-[H-Cu2+-H] in the methyl group model (dP ) CH3). vdW

The E of the methyl group model is larger than that of the hydrogen atom model due to the pseudo-π-electron delocalization in the methyl group model. 3.1.2. Cis and Trans Isomers. We next discuss the difference between the cis-[H-Cu2+-H] and the trans-[H-Cu2+-H] from the view of both the stacking and the total energies. Figure 3 shows the results for the stacking energy. The stacking energy of the cis-[H-Cu2+-H] is almost the same as that of the trans[H-Cu2+-H]. These results indicate that the positions of the nitrogen atoms do not affect stacking energy so much, though the difference between the cis-[H-Cu2+-H] and the trans[H-Cu2+-H] would appear in the total energy as shown below. We next compared the total energy of two base pairs of [H-Cu2+-H], where we fix the r value at 3.60 Å (near the minimum of the interaction energy). Table 1 shows the total energy of one or two [H-Cu2+-H] base pair(s). The trans[H-Cu2+-H] dimer is as stable as the cis-[H-Cu2+-H] dimer in both models. This energy difference is twice as large as one

TABLE 1: Total Energies of cis-[H-Cu2+-H] and trans-[H-Cu2+-H] dimer (in au) with the vdW Correction, Where We Set r Value as 3.60 Åa cis/hyd trans/hyd cis/met trans/met

1 pair

2 pairsb

2 pairs + ALL

-2512.803 079 -2512.803 369 -2591.161 451 -2591.161 711

-5025.595 016 -5025.595 743 -5182.310 952 -5182.311 668

-5025.618 127 -5025.618 840 -5182.337 265 -5182.337 983

Hyd and met mean hydrogen atom model (dP ) H) and methyl group model (dP ) CH3), respectively. b Basis set superposition error was corrected with counterpoise method. a

of the monomers. Therefore, a difference in the cis and trans isomers originates form the stability only of the monomer. This tendency agrees with the fact that there is no difference in the interbase interaction. In what follows, we only show results of the trans isomer rather than the cis isomer to avoid redundancy. 3.2. With and without Cu Ions. In 2002, Tanaka and coworkers synthesized a base pair, [H-2H+-H], which does not

Copper Ion Containing Artificial DNA SCHEME 2: Chemical Structural Formula of [H-2H+-H], Where dP Represents the Backbone Molecule

J. Phys. Chem. B, Vol. 112, No. 51, 2008 16963 TABLE 2: Summary of the Stacking Energy at r ) 3.60 Å (in kcal/mol) [H-Cu2+-H] (r ) 3.60 Å), dP ) CH3 [H-2H+-H] (r ) 3.60 Å), dP ) CH3 2 AT base pairs a

have the copper ion.41 In this section, we compared the trans[H-2H+-H] with the trans-[H-Cu2+-H] in relation to the stacking energy. Hereafter we omit trans for simplicity. [H-2H+-H] is described as Scheme 2. We here adopt the methyl group model (dP ) CH3). [H-2H+-H] is stable because of the two hydrogen bonds similar to those in the adeninethymine (AT) base pair. Therefore, we here consider [H-2H+-H] dimer as the analogue of the AT base-pair dimer. Comparing trans-[H-Cu2+-H] with [H-2H+-H], it is expected that it is possible to observe an origin of the metal-metal interaction. Figure 4 shows the results of the interaction energy of the [H-2H+-H] dimer. According to the figure, the stacking energy of the [H-2H+-H] dimer is estimated to be about 9.3 kcal/mol. In comparison with the previous results, the stacking energy of [H-Cu2+-H] is almost same as that of [H-2H+-H]. The difference in the inter-base-pair distance is negligibly small. As mentioned before, we consider [H-2H+-H] as an analogue of AT pair dimer. Sˇponer computed interaction energy between base pairs with high-level ab initio calculation such as CCSD(T).42,43 It is difficult to evaluate the stacking energy of the present system by the same quality because the system of interest is too large. In order to check the reliability of the present method and understand how much is the error caused by constraint of whole structures, we evaluate the stacking energy of two AT pairs by using the LC-BOP+ALL method. We used the same structure of two AT pairs (which is available in the Supporting Information of their paper), where the r value was fixed at 3.60 Å. Table 2 summarizes the results of the stacking energy. The LC-BOP+ALL method agrees with the result of CCSD(T) for the two AT pairs. The stacking energy of [H-2H+-H] is smaller than that of the natural B-DNA by 4.3 kcal/mol. This energy difference originates from Eint, which can be explained by partial constraint of [H-Cu2+(2H+)-H]. On the other hand, the dispersion correction EvdW is almost the same in all cases. Judging from these results, we have revealed that the chemical origin of the structural stability of [H-Cu2+-H] is similar to that of [H-2H+-H] and of course that of the AT pair, which

Eint

EvdW

total Eint

7.4 7.7 4.1

-16.6 -16.9 -17.6

-9.2 -9.3 -13.6a

The reference value is -14.7 kcal/mol.

TABLE 3: Summary of the Maximum Stacking Energy (Denoted as Estack) (in kcal/mol) and the r Value (in Å) trans-[H-Cu -H] 2+

[H-2H+-H]

dP ) H dP ) CH3 dP ) CH3

r

Estack

3.61 3.57 3.56

7.6 9.2 9.3

is mainly due to the vdW interaction. These results also suggest that the spin-spin interaction among the metal ions does not play a dominant role in forming artificial DNA arrays as discussed below. 3.3. Summary for Staking Energy. Since the ALL functotal tional has r-6 terms, we fitted the interaction energy (Eint ) as total -6 power series of r ; that is, we approximate Eint as 4

total Eint ≈

c

∑ r6ii

(3)

i)1

where ci is an expansion coefficient. Errors between fitted data and calculated values around the local minimum of the interaction energy are within 0.1 kcal/mol so that the fitted data well reproduce the calculated values. We obtained the r value, which gives the minimum of the interaction energy (the maximum stacking energy) from eq 3. Table 3 lists the summary of the stacking energy and the r value. Note that the results for the cis isomer are almost the same as those for the trans isomer, where the deviations from the trans isomer are within 0.1 kcal/ mol for the stacking energy and 0.01 Å for the interbase distance. From the table, it is clear that the r value is around 3.60 Å in all models. Solvent effects make r value longer and stacking energy larger. Our results support both the latest computed results from Jishi (3.59 Å) and experimental data (3.7 ( 0.1 Å). It is noted here that our model has several limitations. For example, we did not consider the effects of a slide of the hydroxypyridone bases and of torsion of the structure. We will consider these effects elsewhere. 4. Singlet-Triplet Energy Gap According to the results of Tanaka et al., the spin state of the copper ions is ferromagnetic. Therefore, the spin state of

Figure 4. Interaction energy between (a) trans-[H-Cu2+-H] (with copper ions) and (b) [H-2H+-H] (without copper ions) in methyl group model (dP ) CH3).

16964 J. Phys. Chem. B, Vol. 112, No. 51, 2008

Matsui et al.

TABLE 4: Energy Difference between Singlet and Triplet of the trans-[H-Cu2+-H] singlet (au) dP ) H dP ) CH3

triplet (au)

LC-BOP -5025.600 450 -5025.600 445 EvdW -0.023 097 -0.023 097 LC-BOP -5182.316 655 -5182.316 650 vdW -0.026 314 -0.026 314 E

difference Jab (K) (cm-1) 1.7 0.0 1.7 0.0

-1.1 0.0 -1.1 0.0

the [H-Cu2+-H] dimer should be the triplet. Although we have assumed that the spin state is the triplet in this study so far, the singlet and triplet states are very close to each other as shown below. In this section, we discuss the energy gap between singlet and triplet with open-shell DFT with the vdW correction and possibility of the spin state observed in the actual experiments. According to the J-model proposed by Yamaguchi and coworkers,44 the spin-spin interaction coupling term Jab can be estimated by following formula

Jab )

Acknowledgment. T.M. and H.M. are thankful to the research Fellowship for young scientist by Japanese Society for the Promotion of the Science (JSPS). This research is supported by a Grant-in-Aid for Global COE Research “Innovation of chemistry” from the Ministry of Education, Science and Culture, and Sports of Japan. This research is also supported by a Grantin-Aid for Core Research for Evolutional Science and Technology from the Japan Science and Technology Agency, by a Grant-in-Aid for the scientific research on priority area “Molecular theory for real systems” (No. 20038008), and Grant-inAid for Young Scientists (B) (No. 20750004). Supporting Information Available: Optimized structures of cis-[H-Cu2+-H] and trans-[H-Cu2+-H]. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes

LS

E-

HS

〈S2〉 -

LS

HS

that the antiferromagnetic (singlet) and ferromagnetic (triplet) states should mix due to the thermal excitation.

E 〈S2〉

(4)

In this case, the whole system is antiferromagnetic when Jab < 0 and is ferromagnetic when Jab > 0. This model reproduces well the experimental data.45 Table 4 lists the energies of the singlet and triplet states for the trans-[H-Cu2+-H]. Again we found that those for the cis isomer are almost the same so that the cis-trans isomerization does not affect the spin state. From the table, the energy differences between the triplet and singlet states are no more than kBT ) 2 (K), and spin-spin interaction is about -1 cm-1 in every case. The vdW correction does not change these tendencies so that it is difficult to find the singlet-triplet energy gap. The modeling of the backbone molecule (dP) is not a problem because the backbone molecules are too far from the copper ions to affect the spin state. These tendencies do not change with the other theoretical method (e.g., Hartree-Fock, MP2, or with other basis sets). Since the singlet-triplet energy gap is quite small, the singlet and triplet states should mix due to thermal excitations. If we assume the Boltzmann distribution of two spin states, there exists the triplet state for [H-Cu2+-H] dimer even at 1.5 K; for example, the distribution of the singlet and triplet species is about 7:3 for kBT ) 1.35 (K), whose triplet state signal may be observed in the actual experiment. Thus, the present calculation does not contradict the experimental evidence, and we propose that the singlet state exists as well as the triplet state. 5. Conclusion According to the results without the vdW correction, it is necessary to consider the weak interaction such as the vdW force in computing the stacking energy as in the case of the natural B-DNA. The distance between the copper ions is about 3.60 Å regardless of the spin state. These results agree well with the experimental data. Similar to the natural B-DNA, the inter-base-pair interaction originates mainly from the vdW force in the artificial DNA. The stacking energy of trans-[H-Cu2+-H] is a little smaller than that of the natural B-DNA and almost the same as that of [H-2H+-H]. This indicates that the copper ions do not affect too much the stability of the artificial DNA. From the results, singlet and triplet are almost degenerated. Our results indicate

(1) Saito, I.; Nakamura, T.; Nakatani, K.; Yoshioka, Y.; Yamaguchi, K.; Sugiyama, H. J. Am. Chem. Soc. 1998, 120, 12686. (2) Giese, B.; Amaudrut, J.; Ko¨hler, A.-K.; Spormann, M.; Wessely, S. Nature (London) 2001, 412, 318. (3) Cai, Z.; Li, X.; Sevilla, M. D. J. Phys. Chem. B 2002, 106, 2755. (4) Basko, D. M.; Conwell, E. M. Phys. ReV. Lett. 2002, 88, 098102. (5) Shimazaki, T.; Asai, Y.; Yamashita, K. J. Phys. Chem. B 2005, 109, 1295. (6) Tsukamoto, T.; Ishikawa, Y.; Natsume, T.; Dedachi, K.; Kurita, N. Chem. Phys. Lett. 2007, 441, 136. (7) Okahata, Y.; Kobayashi, T.; Tanaka, K.; Shimomura, M. J. Am. Chem. Soc. 1998, 120, 6165. (8) Hartwich, G.; Caruana, D. J.; de Lumley-Woodyear, T.; Wu, Y.; Campbell, G. N.; Heller, A. J. Am. Chem. Soc. 1999, 121, 10803. (9) Kelly, S. O.; Boon, E. M.; Barton, J. K.; Jackson, N. M.; Hill, M. G. Nucleic Acids Res. 1999, 27, 4830. (10) Okamoto, A.; Kamei, T.; Tanaka, K.; Saito, I. J. Am. Chem. Soc. 2004, 126, 14732. (11) Tada, T.; Kondo, M.; Yoshizawa, K. ChemPhysChem 2003, 4, 1256. (12) Zilberberg, I. L.; Avdeev, G. M.; Zhidomirov, G. M. J. Mol. Struct.: THEOCHEM 1997, 418, 73. (13) Hush, N. S.; Schamberger, J.; Bacskay, G. B. Coord. Chem. ReV. 2005, 249, 299. (14) Abrescia, N. G. A.; Malinina, L.; Fernandez, L. G.; Huynh-Dinh, T.; Neidle, S.; Subirana, J. A. Nucleic Acids Res. 1999, 27, 1593. (15) Baeyens, K. J.; DeBondt, H.; Pardi, A.; Holbrook, S. R. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 12851. (16) Kankia, B. I. Nucleic Acids Res. 2000, 28, 911. (17) Robinson, H.; Gao, Y.; Sabusgvill, R.; Joachimiak, A.; Wang, A. H. J. Nucleic Acids Res. 2000, 28, 1760. (18) Lee, J. S.; Latimer, L. J. P.; Reid, R. S. Biochem. Cell Biol. 1993, 71, 162. (19) Fuentes-Cabrera, M.; Sumpter, B. G.; Spˇoner, J.; Spˇoner, L.; Petit, L.; Wells, J. C. J. Phys. Chem. B 2007, 111, 870. (20) Miyake, Y.; Togashi, H.; Tashiro, M.; Yamaguchi, H.; Oda, S.; Kudo, M.; Tanaka, Y.; Kondo, Y.; Sawa, R.; Fujimoto, T.; Machinami, T.; Ono, A. J. Am. Chem. Soc. 2006, 128, 2172. (21) Matsui, T.; Shigeta, Y.; Hirao, K. Chem. Phys. Lett. 2006, 423, 331. (22) Matsui, T.; Shigeta, Y.; Hirao, K. J. Phys. Chem. B 2007, 111, 1176. (23) Zhang, Y.; Huang, K. J. Mol. Struct.: THEOCHEM 2007, 812, 51. (24) Voityuk, A. A. J. Phys. Chem. B 2006, 110, 21010. (25) Tanaka, K.; Shionoya, M. J. Org. Chem. 1999, 64, 5002. (26) Meggers, E.; Holland, P. L.; Tolman, W. B.; Romesberg, F. E.; Schultz, P. G. J. Am. Chem. Soc. 2000, 122, 10714. (27) Switzer, C.; Sinha, S.; Kim, P. H.; Heuberger, B. D. Angew. Chem., Int. Ed. 2005, 44, 1529. (28) Tanaka, K.; Yamada, Y.; Shionoya, M. J. Am. Chem. Soc. 2002, 124, 8802. (29) Tanaka, K.; Tengeiji, A.; Kato, T.; Toyama, N.; Shionoya, M. Science 2003, 299, 1212. (30) Tanaka, K.; Shionoya, M. Chem. Lett. 2006, 35, 694. (31) Zhang, H. Y.; Calzolari, A.; Di Feilice, R. J. Phys. Chem. B 2005, 109, 15345. (32) Jishi, R. A.; Bragin, J. J. Phys. Chem. B 2007, 111, 5357.

Copper Ion Containing Artificial DNA (33) Sato, T.; Tsuneda, T.; Hirao, K. J. Chem. Phys. 2005, 123, 104307. (34) Boys, S. F.; Bernardi, F. Mol. Phys. 1970, 19, 553. (35) Andersson, Y.; Langreth, D. C.; Lundqvist, B. I. Phys. ReV. Lett. 1996, 76, 102. (36) Kamiya, M.; Tsuneda, T.; Hirao, K. J. Chem. Phys. 2002, 117, 6010. (37) Sato, T.; Tsuneda, T.; Hirao, K. J. Chem. Phys. 2007, 126, 234114. (38) Iikura, H.; Tsuneda, T.; Yanai, T.; Hirao, K. J. Chem. Phys. 2001, 115, 3540. (39) Song, J.-W.; Hirosawa, T.; Tsuneda, T.; Hirao, K. J. Chem. Phys. 2007, 126, 154105. (40) Frisch, M. J.; et al. Gaussian 03, Revision D.02, Gaussian, Inc., Wallingford, CT, 2004.

J. Phys. Chem. B, Vol. 112, No. 51, 2008 16965 (41) Tanaka, K.; Tengeiji, A.; Kato, T.; Toyama, N.; Shiro, M.; Shionoya, M. J. Am. Chem. Soc. 2002, 124, 12494. (42) Spˇoner, J.; Juree`ka, P.; Marchan, I.; Luque, E. J.; Orozco, M.; Hobza, P. Chem.sEur. J. 2006, 12, 2854. (43) Juree`ka, P.; Spˇoner, J.; E`erny, J.; Hobza, P. Phys. Chem. Chem. Phys. 2006, 8, 1985. (44) Yamaguchi, K.; Takahara, Y.; Fueno, T. Applied Quantum Chemistry; Smith, V. H., et al., Eds.; D. Reidel Pub.: Boston, 1986; p 155. (45) Yamaguchi, K.; Okumura, M.; Maki, J.; Noro, T.; Namimoto, H.; Nakano, M.; Fueno, T.; Nakasuji, K. Chem. Phys. Lett. 1992, 190, 353.

JP8080707