Synthesis, Properties, and NMR Studies of a C8-Phenylguanine

West Virginia University, P.O. Box 9530, Morgantown, West Virginia 26506, ... and Department of Chemistry and Biochemistry, Florida State University,...
0 downloads 0 Views 174KB Size
Chem. Res. Toxicol. 2003, 16, 1385-1394

1385

Synthesis, Properties, and NMR Studies of a C8-Phenylguanine Modified Oligonucleotide that Preferentially Adopts the Z DNA Conformation Peter M. Gannett,*,† Sue Heavner,† Jonathan R. Daft,† Kevin H. Shaughnessy,‡ Jon D. Epperson,§ and Nancy L. Greenbaum§ West Virginia University, P.O. Box 9530, Morgantown, West Virginia 26506, Department of Chemistry, The University of Alabama, Box 870336, Tuscaloosa, Alabama 35487, and Department of Chemistry and Biochemistry, Florida State University, Tallahassee, Florida 32306 Received February 11, 2003

Carcinogenic aryl hydrazines produce C8-arylated purine adducts. The effect of these adducts on DNA conformation and their role in hydrazine carcinogenesis are unknown. Here, we describe a new synthetic route to produce these adducts that is also compatible with the synthesis of the corresponding phosphoramidites needed for oligonucleotide synthesis. Two oligonucleotides were prepared, an unmodified oligonucleotide, d(5′CGCGCGCGCG3′), and a C8-phenylguanine modified oligonucleotide, d(5′CGCGCG*CGCG3′) (G* ) 8-phenylguanine). These oligonucleotides were compared using thermal denaturation, circular dichroism, NMR, and molecular modeling. The phenyl modification destabilizes the B DNA form and stabilizes the Z DNA form such that the B:Z ratio is near one under physiological conditions. In light of recent studies that show a role for Z DNA in gene expression and cell transformation, Z DNA stabilization by C8-arylguanine formation from aryl hydrazines may be relevant to their role in carcinogenesis.

Introduction C8-substituted guanines have been studied in regard to a variety of chemical carcinogens such as AAF1 (1), agents that lead to 8-oxodG formation (2, 3), nitrous oxide (4), or nitric oxide (5). These studies have often focused on the conformational changes as a consequence of these mutations. For example, 8-oxodG forms Watson-Crick base pairs with cytosine but forms a Hoogsteen like base pair with adenine (6). The latter base pair is due to the syn glycosidic bond preference shown by all known C8 substituted guanines and the structural resemblance of the 8-oxodG Hoogsteen edge to thymidine. The mutational consequence of this is 8-oxodG, which is read during replication as a thymidine, produces base transversions (2, 7-9). A common feature of the C8-substituted guanines noted is that the heteroatom attached to the C8 carbon can participate in hydrogen bonding or exist in various tautomers. Groups pendant to the heteroatom may further interact with the DNA helix, such as AAF, which can intercalate into the helix (10). C8-guanine adducts * To whom correspondence should be addressed. Tel: (304)293-1480. Fax: (304)293-2576. E-mail: [email protected]. † West Virginia University. ‡ The University of Alabama. § Florida State University. 1 Abbreviations: 2-AF, 2-aminofluorene; AAF, N-acetylaminofluorene; CD, circular dichroism; CG, d(CGCGCGCGCG)2; CG8Ph, d(CGCGCG8PhCGCG)2; DMTr, 4,4′-dimethoxytrityl; DMAP, 4-N,N-(dimethylamino)pyridine; DMF, dimethylformamide; FPLC, fast protein liquid chromatography; LAH, lithium aluminum hydride; C5Me, 5-methylcytidine; NBS, N-bromosuccinimide; NOE, nuclear Overhauser effect; 8-oxodG, 8-oxo-2′-deoxyguanosine; PME, particle mesh Ewald; THF, tetrahydrofuran; Tm, thermal denaturation temperature; TPPTS, tris(m-sulfonatophenyl)phosphine trisodium; TEA, triethylamine.

Figure 1. General structure of carcinogenic aryl hydrazines. The aryl hydrazines X ) -CH3, -CH2OH, and -COOH are found in the mushroom of commerce, A. bisporus.

connected by a carbon bond have received less attention, although they have been shown to form from a variety of carcinogens including alkylhydrazines (11), aryl hydrazines (12-15), polyaromatic hydrocarbons (16), diazoquinones (17), alcohols (18, 19), and peroxides (17, 20). Other carcinogens may also lead to the formation of adducts of this nature including azo dyes (21, 22), triazenes (23), and hydrazine-containing pharmaceuticals (24, 25), based on the metabolic intermediates that form from these compounds (26). The formation of C8-alkyl and C8-arylguanine adducts and their relationship to mutagenesis is poorly understood. Without the C8-heteroatom on guanine, it is not obvious how base pairing properties might be altered. Moreover, whether these adducts cause significant conformational changes has only been studied for a few DNA examples. Our interest in this area concerns the C8arylguanine adducts that we have shown form in DNA from carcinogenic aryl hydrazines (Figure 1) (13). Several of these aryl hydrazines (X ) -CH3, -CH2OH, and -COOH; Figure 1) are found in the mushroom of commerce, Agaricus bisporus (27), and are metabolized to arenediazonium ions (28-30) and then to aryl radicals (31) that subsequently lead to the formation of C8arylguanine adducts (12).

10.1021/tx034023d CCC: $25.00 © 2003 American Chemical Society Published on Web 09/17/2003

1386 Chem. Res. Toxicol., Vol. 16, No. 10, 2003

The formation of the C8-arylguanine adducts correlates with the carcinogenicity of the parent arylhydrazine thereby suggesting a role for them in arylhydrazine carcinogensis (28). However, studies of a related adduct, 8-phenylguanine, suggest that C8-arylguanine adducts may be poorly read, although when they are read, they are usually read correctly (32). Thus, the low level of mutation produced in daughter strands suggests that if C8-arylguanine formation is related to carcinogenicity, it is likely the result of an effect other than misreading. In a study of C8-methylguanine, formed from methyl hydrazine, it was shown that the B DNA form of an alternating purine-pyrimidine sequence was destabilized. At sodium chloride concentrations above 5 mM, the Z DNA form predominates (33). Recently, it was shown that Z DNA may regulate gene expression and cell transformation (34). These observations suggested to us the possibility that the C8-arylguanine adducts that we have observed (13, 28) may destabilize B DNA and therefore stabilize the Z DNA conformation when formed in purine-pyrimidine sequences. Here, we have prepared unmodified and modified oligodeoxynucleotides that contain a C8-phenylguanine adduct. These oligonucleotides have been examined by thermal denaturation measurements, CD spectroscopy, and NMR spectroscopy. Our measurements show that under near physiological conditions, the C8-phenylguanine modified oligonucleotide exists in approximately a 1:1 B/Z DNA ratio. Thus, the C8-phenylguanine derivative behaves like the C8-methylguanine adduct and stabilizes the Z DNA form.

Experimental Procedures General Procedures. Solvents and reagents were obtained from Aldrich (Milwaukee, WI) and were used without purification unless otherwise noted. NMR spectra were obtained on a Varian Gemini 300, a Bruker 360, or a Varian Unity 500 spectrometer. Methylene chloride was dried by distillation from phosphorus pentoxide. TEA, pyridine, and THF were dried by distillation from LAH. DMF was purified by distillation from barium oxide. Unmodified phosphoramidite DNA bases and CPG resins were obtained from Glen Research (Sterling, VA). MS were recorded on an Agilent 5973N (low resolution) or a Finnigan MAT 90 (high resolution). Synthesis. 1. 8-Bromo-2′deoxyguanosine (3). 2′-Deoxyguanosine (2.67 g, 0.94 mmol) was suspended in water (300 mL), and NBS (2.68 g, 1.50 mmol) was added. The reaction was stirred for 20 min and filtered. The filter cake was then dried in vacuo to yield 3 (1.64 g, 0.48 mmol, 48% yield) (35). 1H NMR (300 MHz, DMSO-d6): δ ppm 10.82 (1H, s(br), NH), 6.51 (2H, s(br), NH2) 6.16 (1H, dd, J ) 7.2 Hz, H-1′), 4.40 (1H, m, H-3′), 3.81 (1H, m, H-4′), 3.63 and 3.50 (2H, m, H-5′/5′′), 3.17 (1H, p, J ) 6.87, H-2′′), 2.11 (1H, ddd, J ) 2.81, 6.62 Hz, 13.12 Hz, H-2′). 13C NMR (75 MHz, DMSO-d6): δ ppm 155.4, 153.3, 152.0, 120.5, 117.5, 87.9, 85.0, 71.0, 62.0, 36.4. 2. 8-Phenyl-2′deoxyguanosine (4). Palladium acetate (2.2 mg, 0.01 mmol), TPPTS (14.8 mg, 0.025 mmol), sodium carbonate (80 mg, 0.75 mmol), 3 (130 mg, 0.37 mmol), and phenylboronic acid (68.2 mg, 0.56 mmol) were placed in a round bottom flask under nitrogen. Degassed 2:1 water:acetonitrile was added (3.5 mL), and the reaction was heated in an oil bath at 80 °C until RP-TLC (C-18, 1:1 water:methanol) showed complete conversion to the product 4 (2-4 h). The reaction mixture was then diluted with water (20 mL), and the pH was adjusted to 6-7 with 10% aqueous KH2PO4. The mixture was heated to dissolve precipitated solids and then allowed to cool to 0 °C over several hours. The product 4 was isolated by filtration as a cream-colored solid (100 mg, 78%). 1H NMR (360 MHZ, DMSOd6): δ ppm 10.8 (1H, brs, -H), 7.66-7.63 (2H, m, phenyl), 7.57-

Gannett et al. 7.52 (3H, m, phenyl), 6.41 (2H, brs, NH2), 6.07 (1H, dd, J ) 6.38, 8.20 Hz, H-1′), 5.13 (1H, d, J ) 4.46 Hz, 3′-OH), 4.97 (1H, t, J ) 6.38, 5.47 Hz, 5′-OH), 3.79-3.77 (1H, m, H-4′), 3.693.62 (1H, m, H-5′/5′′), 3.57-3.50 (1H, m, H-5′), 3.27-3.13 (1H, m, H-2′), 2.03 (1H, ddd, J ) 1.73, 7.36, 12.97 Hz, H-2′). 13C NMR (90.6 MHz, DMSO-d6): δ ppm 159.8, 155.1, 152.1, 146.3, 130.6, 129.2, 129.1, 128.6, 117.3, 88.1, 85.0, 71.4, 62.3, 36.9. 3. N2-Isobutyryl-8-phenyl-2′-deoxyguanosine (5). Compound 4 (1.15 g, 3.13 mmol) was suspended in dry pyridine (40 mL) and cooled in an ice bath. Chlorotrimethylsilane (5.2 mL, 41.2 mmol) was then added, and the mixture was stirred at 0 °C for 30 min. Isobutyric anhydride (4.8 mL, 29 mmol) was then added, and the mixture was stirred at room temperature for 2 h. After the mixture was cooled at 0 °C in an ice bath, ammonium hydroxide (15%, 32 mL) was added and the mixture was stirred for 30 min. The solvent was removed by evaporation, and the residual solids were suspended in water (40 mL). The water layer was washed with diethyl ether (3 × 100 mL). The ivory precipitate appearing in the aqueous phase was collected and dried in vacuo to yield 5 (1.05 g, 2.54 mmol, 81%). 1H NMR (300 MHz, DMSO-d6): δ 12.20(1H, s, NH), 11.46(1H, s, NH2), 7.72 (2H, m, phenyl), 7.57 (3H, m, phenyl), 6.14 (1H, dd, J ) 7.20 Hz, H-1′), 5.17 (1H, d, J ) 4.2 Hz, H-3′-OH), 4.78 (1H, m, 5′-OH), 4.35 (1H, m, H-3′), 3.78 (1H, m, H-4′), 3.34-3.25 (1H, m, H-5′′), 3.25-3.23 (1H, m, H-5′), 3.20 (1H, m, H-2′′), 2.82 (1h, p, J ) 6.90 Hz, -CH(CH3)2), 2.08 (1H, m, H-2′), 1.14 (6H, d, J ) 6.60 Hz, -CH(CH3)2). 4. 5′-O-(DMTr)-N2-isobutyryl-8-phenyl-2′-deoxyguanosine (6). Compound 5 (180 mg, 0.44 mmol) was dried in vacuo under P2O5, and then, dry pyridine (4 mL) and TEA (85 µL, 0.61 mmol) were added. DMTr-Cl (216 mg, 0.60 mmol) was then added to the reaction flask and stirred for 2.5 h under nitrogen at room temperature. Methanol (3 mL) was added to quench the reaction, and the solvent was removed in vacuo. The product was purified by low-pressure column chromatography (silica gel, 49:1 CH2Cl2:CH3OH) to yield 6 (179 mg, 57%). 1H NMR (300 MHz, DMSO-d6): δ ppm 12.10 (1H, brs, NH) 11.20 (1H, brs, NH), 7.75 (2H, m, phenyl), 7.56 (3H, m, phenyl), 7.34-6.65 (13H, m, DMTr-H), 6.22 (1H, m, 1′-H), 5.14 (1H, m, 3′-OH), 4.44 (1H, m, 3′-H), 3.98 (1H, m, 4′-H), 3.70 and 3.68 (3H each, s, -OCH3), 3.50/3.07 (2H, m, H-5′/5′′), 3.20 (1H, m, 2′′), 2.77 (1H, m, -CH(CH3)2), 2.20 (1H, m, 2′), 1.13 (J ) 6.60, 6H, d, CH(CH3)2). 5. 3′-O-[(2-Cyanoethoxy)(diisopropylamino)phosphino]5′-O-(DMTr)-N2-isobutyryl-8-phenyl-2′-deoxyguanosine (7). Compound 6 (137 mg, 0.19 mmol) was dried over P2O5 and dissolved in CH2Cl2 (2 mL). TEA (36 µL, 0.26 mmol) and then 2-cyanoethyl N,N-diisopropylchlorophosphoramidite (36 ul, 0.16 mmol) were added, and the mixture was stirred for 30 min at room temperature. A second portion of 2-cyanoethyl-N,Ndiisopropylchlorophosphoramidite (20 µL, 0.09 mmol) was then added, and stirring was continued for an additional 30 min. The solvent was removed in vacuo, the residue was resuspended in 4:1 benzene:THF (4 mL) and filtered, and the filtrate was concentrated in vacuo. The product was purified by low-pressure column chromatography (silica, 3:1 ethyl acetate:hexanes). Fractions containing 7 were pooled and concentrated in vacuo. The product was then coevaporated with benzene (3 × 5 mL) to yield 7 (76 mg, 0.08 mmol, 43%). 1H NMR (300 MHZ, DMSOd6): δ ppm 12.00 (1H, brs, NH) 11.01 (1H, brs, NH), 7.55 (2H, m, phenyl), 7.42 (1H, dd, J ) 2, 8.5 Hz, ArH-4′′), 7.35 (3H, m, phenyl), 7.26-7.39 (4H, m, ArH-2′′,3′′,5′′,6′′), 7.26 (4H, d, J ) 8, ArH-2,2′,6,6′), 6.85 (4H, d, J ) 8, ArH-3,3′,5,5′), 6.10 (1H, m, 1′-H), 4.22 (1H, m, 3′-H), 3.92 (1H, m, POCH), 3.90 and 3.88 (3H each, s, -OCH3), 3.81 (1H, m, 4′-H), 3.74 (1H, m, POCH), 3.67/3.71 (1H, m, CH(CH3)2), 3.35/3.01 (2H, m, H-5′/5′′), 3.02 (1H, m, 2′′), 2.63/2.73 (2H, t, J ) 6 Hz, CH2CN), 2.55 (1H, m, -{CH}(CH3)2), 2.10 (1H, m, 2′), 1.250/1.251 (12H, s, CH(CH3)2), 1.05 (6H, d, J ) 6.60, CH(CH3)2). Oligonucleotide Synthesis. Large scale (10-20 mmol) synthesis was conducted on modified ABI 430A protein synthesizer. The oligonucleotides, CGCGCGCGCG (CG) and CGCGCG8-PhCGCG (CG8-Ph), used in this study were synthe-

C8-Phenylguanine Modified Oligonucleotides

Chem. Res. Toxicol., Vol. 16, No. 10, 2003 1387

Table 1. Melting Temperatures for CG and CG8Ph Tm (°C) NaCl (mM)

CGa

CG8Ph

25 100 4500

64 65 66

64b 65b 66

a All samples contained approximately 0.5 µM oligonucleotide dissolved in 10 mM phosphate buffer, pH 7.4. Data were collected from 20 to 95 °C at a heating rate of 0.25°/min. b At the indicated salt concentration, CG8-Ph was not present in a single form (see text).

sized using the solid phase phosphoramidite protocol. The synthetic steps were detritylation, coupling, capping, and oxidation. The oligonucleotides (10-20 µM scale) were cleaved off of the resin by treatment with concentrated NH4OH (28-30%, 12 mL) at room temperature for 1 h and filtered through a 0.2 µm filter disk. The cleavage of protecting groups was accomplished by heating the filtrate at 55 °C for 20 h and then dried down on a SpeedVac. Final purification of oligomers was achieved by FPLC using a Bio-Rad TSK DEAE-5-PW column. Both oligonucleotides were purified using a linear gradient (10-70% B over 60 min; buffer A: 10 mM NaOH, pH 11.8; buffer B: 10 mM NaOH, 1 mM NaCl, pH 11.8; flow rate, 7 mL/min; UV detection, 260 nm). The oligomers were desalted with reverse phase Waters Sep-Pack (C-18) cartridges by 60% MeOH/water elution. The purified oligonucleotides were examined by mass spectrometry (MALDI-TOF) and were shown to be the desired products. The efficiency of the coupling of 7 was typically in the range of 98-99%. Thermal Denaturation. All DNA samples were made up in phosphate buffer (10 mM NaH2PO4, pH 7.4) and varying concentrations of sodium chloride, heated to 90 °C for 30 min, and then cooled slowly to room temperature. UV-monitored melting temperature experiments were conducted by monitoring at 260 nm using a Cary 300 spectrometer on the duplexes (CG) and CG8-Ph, under the following conditions: 10 mM sodium phosphate buffer, pH 7.4, and at varying NaCl concentrations (Table 1). Spectra were recorded over the temperature range of 20-90 °C and at a rate of 0.25 °C/min. CD. The CD spectra were recorded on an AVIV model 62A CD spectrometer. Solution concentrations of each oligonucleotide were approximately 25 µM. Unless otherwise stated, solutions were prepared in 10 mM phosphate buffer, pH 7.4, and varying amounts of NaCl were added (see figure captions) and DNA strands annealed as described for the thermal denaturation samples. The final sample volume was 400 µL. NMR. The NMR samples were made up in 10 mM phosphate buffer, pH 7.4, with 25 mM, 100 mM, 200 mM, and 1 M NaCl in D2O (1 mM in duplex). All NMR spectra were measured on a Varian Unity 500 MHz spectrometer at 28 °C. The twodimensional NOESY spectra of nonexchangeable protons were collected at 28 °C with a mixing time of 150 ms. The data were collected with 512 t1 increments and 2048 t2 complex points, each the sum of 64 transients. Data were apodized with shifted sine bell functions in both dimensions and zero filled to give a 4k by 1k data set. The two-dimensional NOESY spectra of exchangeable protons were acquired on samples at 28 °C in 10% D2O/90% H2O solution using the 1-1 pulse sequence, with a mixing time of 150 ms and a sum of 64 transients. Molecular Modeling. Molecular modeling and dynamics were performed with Amber and the Cornell 95 force field (36). Because atom parameters for C8-phenylguanine were not available, these were developed. The modified base was built and optimized in Gaussian98 (HF 6-31G*, B3LYP basis set) to provide optimum bond lengths, angles, atom charges, and the torsional profile. These parameters where then adjusted for the Cornell 95 force field using the Amber program. The final force field parameters and how they were developed are provided in the Supporting Information. All oligonucleotide structures were built using Sybyl (Tripos, Inc.) and then transferred to Amber 6. The unmodified se-

quences (B and Z) were 5′CGCGCGCGCG3′ (CG). The modified sequences were constructed from the unmodified sequences by attaching a phenyl group on the C8-position of G6 and G16. Hydrogen atoms, 18 sodium counterions to neutralize the charge, and a water box were added within xleap. Initially, the water and solute were equilibrated by minimizing the water and counterions with the DNA fixed (1000 steps) followed by 25 ps of nonPME dynamics, raising the tempurature from 100 to 300 K (DNA fixed), then 25 ps of PME dynamics to allow the watercounterion system to equilibrate, minimization (1000 steps), and then 3 ps of dynamics. This was followed by five consecutive 600 step minimizations, decreasing the harmonic potential from 20 to 0 kcal/mol in 5 kcal/mol steps, a final equilibration of 10 ps PME dynamics with no restraints on DNA, counterions, or water was performed during which the system was heated from 100 to 300 K. Production dynamics runs were then conducted for approximately 2 ns. The “most representative” structure for each conformer was generated using the cluster trajectory option in the Dynamics Menu of MOIL-view (37). Trajectories were split into 250 frame pieces using the ptraj module of Amber, necessary due to size limitations in MOIL-view for this particular trajectory analysis tool. The steps performed in the Cluster analysis program were as follows: (i) each frame is initially placed in a cluster by itself; (ii) a cutoff distance of 2 Å is specified; (iii) on the basis of a two-dimensional RMSD matrix generated, the average RMSD between all pairs of structures for each pair of clusters was calculated; (iv) the average RMSD for the most similar of these pairs is compared to the specified cutoff value; and (v) if the value is less than the cutoff value, the two clusters are combined. This process is repeated until all cluster pairs have an average RMSD greater than the cutoff value. The most representative structure of each cluster is that structure that has the lowest average RMSD to all other members of it’s cluster.

Results Phosphoramidite Synthesis. Initially, the method used to prepare the C8-phenylguanine modified phosphoramidite (7) was essentially the same as that reported by Kohda et al. (32). However, the initial step in this synthetic scheme, phenylation by base-induced decomposition of benzene diazonium ion, is difficult to control, poor yields were obtained, and the product was difficult to purify. In addition, while this reaction works for simple substrates, it is not certain that more complex substrates would be able to survive the reaction conditions. Therefore, we have developed an alternative and more general route to C8-arylguanine derivatives. The new method developed was based on the Suzuki coupling reaction between aryl halides and boronic acids (38). The reaction is catalyzed by a combination of Pd(0) and a tertiary phosphine ligand that ensures catalyst solubility and stability. This methodology has recently been applied to the arylation of 2′-deoxyguanosine derivatives (39, 40). Typically, the Suzuki coupling is carried out in organic solvents, which in the case of 2 requires its conversion to more hydrophobic forms, such as the 3′,5′-bis-(tert-butyldimethylsilyl)ether. After arylation, the protecting groups would have to be removed adding at least two steps to the overall synthetic scheme. To avoid a protection/deprotection sequence, we have explored the use of water soluble phosphine ligands that would allow the reaction to be conducted in an aqueous environment (41, 42). The new synthetic scheme is shown in Figure 2 and begins with the reaction of 2 and NBS to form 3 (35). The next step is the aqueous phase Suzuki coupling of 3 and phenylboronic acid catalyzed by palladium acetate

1388 Chem. Res. Toxicol., Vol. 16, No. 10, 2003

Gannett et al.

Figure 2. Synthetic scheme for the synthesis of the phosphoramidite of C8-phenyl-2′-deoxyguanosine. Conditions: (a) NBS, water, 45 min; (b) phenylboronic acid, TPPTS, Na2CO3, Pd(OAc)2, 1:2 MeCN/H2O, 80 °C; (c) i-butyryl anhydride, pyridine; (d) DMTr-Cl, pyridine, DMAP, TEA; and (e) (N(i-Pr)2)P(Cl)(O(CH2)2CN), TEA, CH2Cl2.

(2.5 mol %) and TPPTS (6.25 mol %) as the water soluble ligand. The reaction is conducted in a mixed solvent system of water and acetonitrile under basic conditions (sodium carbonate) (42). This method appears to be quite general, and we have used it to prepare several other C8aryl substituted guanosines in addition to the desired 8-phenyl-2′-deoxyguanosine, 4, described here. Yields of the arylated products are typically 80% or higher. Following the coupling reaction, 4 was converted to the N2-isobutyrylamide 5, then the 5′-dimethoxytrityl ether 6, and finally the phosphoramidite 7. The only difficulty encountered during the course of these three standard reactions was with the dimethoxytritylation reaction. We found that 6 was quite sensitive to acid and work up of the dimethoxytritylation reaction required some care so as not to expose it to acidic conditions. Once isolated, 6 appeared to be quite stable. This method has been used with several substituted phenyl boronic acid derivatives and with 8-bromo-2′-deoxyadenosine in place of 3. C8arylated products are typically obtained in 50-80% yields. These should be readily converted to the corresponding phosphoramidites. Therefore, using this methodology, it should be possible to prepare a wide range of C8-arylpurine modified oligonucleotides. Thermal Denaturation. Thermal denaturation curves were obtained for solutions of CG and CG8Ph at low, intermediate, and high salt concentrations. At low salt concentrations (25 mM), where the B form is expected to predominate for both oligonucleotides, the unmodified 10-mer CG displayed a well-defined melting curve. Similar Tm curves were also observed for CG and CG8Ph at high sodium chloride concentration (4.5 M). The Tm values are shown in Table 1. The unmodified oligonucleotide CG and modified oligonucleotide CG8Ph gave similar curves and Tm temperatures at 25 and 100 mM sodium chloride. However, as shown by the CD and NMR data, the CG8Ph oligonucleotide exists as a mixture of B and Z DNA, at these salt concentrations, and thus interpretation of the Tm data is not straightforward. CD Spectroscopy. CD spectroscopy can distinguish between B and Z DNA. The CD signal of B DNA shows

positive ellipticity at approximately 280 and 220 nm and a negative ellipticity at approximately 248 nm. In contrast, Z DNA shows a negative ellipticity at approximately 295 and 248 nm (33, 43, 44). In Figure 3 is shown the CD spectra of CG and CG8Ph as a function of salt concentration. The salt concentrations were varied from 0 to 4600 mM for CG and between 0 and 3600 mM for CG8Ph. The CG oligonucleotide at salt concentrations below 2 M is predominantly in the B form, as indicated by the positive ellipticity at 280 nm. Above 2 M salt, the Z form predominates. Overall, these observations for the unmodified oligonucleotide CG are consistent with what has been reported for related oligonucleotides (33). The CD spectra of the modified oligonucleotide, CG8Ph, are shown in Figure 3b. The dependence of this spectrum on salt concentration is remarkably different with respect to CG. First, at low salt concentration, the 248 nm minimum is attenuated relative to the unmodified oligonucleotide. Second, by inspection of the spectrum between 280 and 295 nm, it can be seen that roughly equivalent amounts of the B and Z DNA forms are present at salt concentrations between 200 and 400 mM sodium chloride, 5-10 times less than required for CG. The CD data for CG8Ph are similar to other related oligonucleotides that have alternating purine/pyrimidine sequences and a C8-modified guanine. Many of these studies were conducted with poly(dG-dC) that had been treated with agents known to form C8 adducts such as bromine (43), N-methyl-4-aminoazobenzene (45), or 2-AF (46). These systems usually require salt concentrations in the range of 1-2 M sodium chloride to produce a predominance of the Z DNA conformation. For example, at 11% incorporation of 2-AF in poly d(G-C), salt concentrations above 1.2 M are required for the Z DNA form to predominate. NMR. Characteristic through-space and through-bond interactions observed as cross-peaks in NOESY or COSY spectra, respectively, were used to help distinguish conformations of the duplexes. COSY and NOESY data were obtained on samples made up in buffer with low salt concentration (25 mM NaCl) and were assigned by

C8-Phenylguanine Modified Oligonucleotides

Chem. Res. Toxicol., Vol. 16, No. 10, 2003 1389 Table 3. Chemical Shifts of CG8Ph in the Z DNA Conformationa base no. C1 G2 C3 G4 C5 G68Ph C7 G8 C9 G10

H5/Ph H8/6 5.90

7.58 7.93 5.24 7.49 7.91 5.34 7.56 7.82, 7.67, 7.51 5.24 7.57 7.93 5.36 7.55 8.03

H1′

H2′

H2′′

H3′

H4′

H5′

H5′′

5.86 6.35 5.85 6.33 5.91 6.4

1.77 2.87 1.82 2.82 1.87 2.81

2.55 2.87 2.74 2.82 2.76 2.81

4.73 5.12 4.93 4.95 4.94 4.96

3.78 4.28 3.89 4.24 3.90 4.41

2.72 4.17 2.72 4.13 2.71 4.23

3.19 4.05 3.90 4.01 3.90 4.02

5.99 6.36 5.87 6.39

1.86 2.88 1.82 3.33

2.78 2.88 2.79 2.57

5.10 5.13 4.94 4.96

3.90 4.29 3.88 4.33

2.72 4.18 2.72 4.18

3.74 4.18 3.88 4.15

a Sample conditions were 1 mM CG8Ph, 10 mM phosphate buffer (pD 7.4), and 1 M sodium chloride; temperature, 28 °C.

Figure 3. CD spectra of (a) CG and (b) CG8Ph as a function of salt concentration. All spectra were recorded at 20 °C in solutions containing approximately 25 µM oligonucleotide and 10 mM phosphate buffer, pH 7.4. Successive concentrations of salt are (a) 0, 50, 200, 400, 600, 2000, 3000, and 4500 mM NaCl and (b) 0, 2, 4, 25, 50, 100, 200, 400, 600, 2000, and 3600 mM NaCl. Table 2. Chemical Shifts of CG in the B DNA Conformationa base no. C1 G2 C3 G4 C5 G6 C7 G8 C9 G10

H5

H8/6

H1′

H2′

H2′′

H3′

H4′

H5′

H5′′

5.88

7.61 8.01 7.35 7.93 7.33 7.92 7.34 7.93 7.36 7.95

5.78 5.94 5.77 5.91 5.72 5.90 5.73 5.92 5.85 6.17

1.95 2.65 1.94 2.66 2.00 2.66 1.92 2.70 1.94 2.42

2.41 2.8 2.41 2.75 2.42 2.74 2.35 2.76 2.36 2.66

4.88 5.02 4.88 5.02 4.88 5.02 4.85 5.02 4.84 4.72

4.20 4.40 4.16 4.39 4.23 4.40 4.20 4.40 4.20 4.15

4.16 4.14 4.13 4.14 4.16 4.14 4.13 4.14 4.16 4.10

4.16 4.06 4.13 4.06 4.16 4.06 4.13 4.06 4.16 3.75

5.45 5.41 5.41 5.48

a Sample conditions were 1 mM CG, 10 mM phosphate buffer (pD 7.4), and 100 mM sodium chloride; temperature, 28 °C.

standard procedures. We first acquired NMR spectra of the unmodified oligonucleotide (CG) to serve as a reference spectrum for the B DNA form. Chemical shift assignments are shown in Table 2. Two-dimensional NOESY spectra of exchangeable protons displayed minimal dispersion of resonances belonging to imino and amino protons because of the repeated base sequence (47). As a result, only two distinct imino proton resonances (at 12.99 and 13.02 ppm) were observable. Likewise, only two resonances attributable to base-paired amino protons (8.35 and 8.37 ppm) and nonbase-paired amino protons (6.42 and 6.46 ppm) were detected. Sequential assignments were made chiefly from NOESY

spectra in D2O. The pattern of intra- and interresidue NOEs was consistent with the oligonucleotide adopting a B DNA structure. Similar measurements were then made at high salt, which favors Z DNA formation. However, we were unable to obtain clear NMR spectra of CG in the Z DNA form because the high salt concentrations required to convert CG completely into the Z DNA form (4.5 M NaCl) produced significant line broadening. NMR spectra of the 8-phenyl modified oligonucleotide CG8Ph were acquired at 25 mM, 100 mM, 200 mM, and 1 M sodium chloride. The chemical shift assignments, determined at 1 M NaCl, are shown in Table 3. In particular, strong intranucleotide G2-H1′/G2-H8, G4-H1′/ G4-H8, G8-H1′/G8-H8, and G10-H1′/G10-H8 cross-peaks were observed, consistent with a Z DNA conformation (48). Additionally, deoxyguanosine residues adopted a syn conformation, identified by intranucleotide H8-H1′ NOE cross-peaks, and the cytosines adopted the anti conformation about the glycosidic bond. The assignments of the sugar-H2′/H2′′ were made using the NOE correlations from the base protons to these sugar protons. All C-H2′ protons in CG8Ph were upfield of the corresponding resonance of the unmodified oligonucleotide (B DNA), and the C-H2′′ and G-H2′/H2′′ protons were all downfield. In the case of the C-H2′′ protons, exclusive of the terminal bases, the observed downfield shift (Z vs B) was substantial (0.3-0.4 ppm range). Correlations between the G-H8/H1′ and the C-H4′/H5′/H5′′ protons were observed, as expected for a Z DNA, and used to assign the latter resonances. The shifts determined for C-H4′/H5′/H5′′ in CG8Ph were all upfield of the corresponding protons in the unmodified oligonucleotide (C-H4′ protons roughly 0.4 ppm upfield shifted, C-H5′ approximately 0.2 ppm upfield, and C-H5′′ greater than 1.4 ppm). Assignments of the G-H4′/H5′/H5′′ protons were based on NOESY correlations between G-H2′ and G-H4′, G-H3′ and G-H5′/H5′′, and COSY correlations from G-H2" f G-H3′ f G-H4′ f G-H5′/H5′′. Thus, unlike CG, it was possible in the case of CG8Ph to assign the G-H3′/H4′/H5′/H5′′ and C-H3′/H4′/H5′/H5′′ protons unequivocally because of their upfield shifts. Finally, as in the case of CG, the imino and amino protons were not resolved; only two distinct imino protons were observable (12.34 and 13.18 ppm). Likewise, only two distinct base-paired amino protons (8.47 and 8.32 ppm) and nonbase-paired amino protons (6.80 and 6.65) were observable. The spectra recorded at lower salt concentrations (25, 100, or 200 mM NaCl) were quite complex, making full

1390 Chem. Res. Toxicol., Vol. 16, No. 10, 2003

Gannett et al.

Figure 4. Aromatic region of the COSY spectrum of CG8Ph showing the phenyl protons in the single-stranded (ss), B DNA (b) and Z DNA (z) forms. The sample included 100 mM NaCl and 10 mM sodium phosphate buffer, pH 7.4, at 20 °C. Table 4. Single-Stranded, B, and Z DNA Chemical Shifts of the G8-Phenyl Protons of CG8Pha single-stranded B DNA Z DNA

H-2/6 (ortho)

H-3/5 (meta)

H-4 (para)

7.78 6.35/7.28 7.84

7.32 7.04/7.15 7.55

7.33 7.17 7.65

a Sample conditions were 1 mM CG8Ph, 10 mM phosphate buffer (pD 7.4), and 100 mM sodium chloride; temperature, 28 °C.

assignment problematic. We tentatively attributed the complexity to the presence of all three forms of CG8Ph (single-stranded, B, and Z forms) in the sample and attempted to identify resonances for phenyl protons associated with each form. Assignment of the phenyl protons in the Z form of CG8Ph was made on the basis of their observed chemical shifts in 1 M NaCl. At the lower salt concentrations, in addition to these previously recorded shifts, additional resonances attributable to phenyl group protons were observed (Figure 4). The additional sets of phenyl resonance have been assigned to the B and a third form that may be the single-stranded form (Table 4). The chemical shifts of the phenyl group in the singlestranded form were similar to the Z DNA form, although they are shifted upfield a small amount. The chemical

shifts for the phenyl group in the B form display a chemical shift range of nearly 1 ppm. As indicated in Table 4, the chemical shift of the phenyl-H2 proton is far upfield at 6.35 ppm (Figure 4). The upfield shift of the phenyl-H2 is substantial, appearing approximately 1.5 ppm upfield of the same proton in the Z form, and is likely due to the location of this proton over C5 (Figure 5a). The phenyl-H3 is also shielded by C5, although not to the same extent. Likewise, the phenyl-H4, H5, and H6 are less shielded by C5 than the phenyl-H2 proton but are still upfield of the corresponding resonances in either the single-stranded or the Z forms of CG8Ph. Finally, a variable temperature experiment was conducted and the phenyl resonances associated to each of the forms were monitored over the temperature range of 10-50 °C. At the lower temperature, the resonances ascribed to the Z form predominated. As the temperature was increased, the B form increased fairly rapidly and the resonances associated to the single stranded form increased more slowly. At 50 °C, the B form predominated. This behavior is in agreement with the assignments based on similar studies conducted with C8methylguanine modified oligonucleotides (33). It was not possible to quantitate accurately the relative amounts

C8-Phenylguanine Modified Oligonucleotides

Figure 5. Partial structures of CG8Ph in the (a) B DNA conformation and (b) Z DNA conformation. Only bases C5(C16), G6(C15), and C7(G14) are shown.

of the B, Z, and single-stranded forms due to overlap of the latter with other resonances (C8H of guanines of all three forms). However, the increase in the B form is unequivocal, as most of the phenyl resonances associated with the B form do not overlap other resonances. Molecular Models. Molecular models of the modified B and Z oligonucleotides were built in Amber and subjected to energy minimization followed by 2 ns of molecular dynamics. A most representative structure for each conformer was then generated by finding the structure with the smallest RMSD as compared with all of the other structures in the dynamics trajectory (37). Partial structures of the resulting B and Z DNA forms of CG8Ph are shown in Figure 5a,b, respectively. All four structures (unmodified and modified, B and Z) demonstrated structural and energetic stability over the course of the 2 ns production simulation. The energy of the unmodified Z DNA structure indicated that it was less favorable than the unmodified B DNA structure; however, 8-phenyl modification of G6 and G16 reversed this and the Z DNA form was favored over the B DNA form. Examination of the modified DNAs suggests an explanation for this result since, in the B DNA conformation, the phenyl ring is in a sterically unfavorable position relative to its position in the Z DNA conformation. The components of the energy were calculated from the trajectories using DELPI (49) in the MM_PBSA module of Amber (50, 51) from the trajectories. The major contributor to the stabilization of the Z forms was the solvation energies. The solvation energies of the Z forms, however, were attenuated by their unfavorable electrostatic repulsion. The van der Waals energies between B and modified B were comparable. However, the internal energy term (bond lengths, bond angles, and dihedral angles) did favor the unmodified B form relative to the

Chem. Res. Toxicol., Vol. 16, No. 10, 2003 1391

modified B form. Finally, comparison of the helical parameters for the unmodified and modified B DNAs showed the sugar pucker for the modified bases to be significantly different than that of the unmodified bases. This was not the case for the Z DNAs in which the sugar pucker for all bases was within the range typical for Z DNA. The B DNA form of CG8Ph (Figure 5a) shows that the phenyl ring lies in the major groove with the phenyl-H2 and H3 protons over C5. In this conformation, the five phenyl protons are expected to all be nonequivalent, in agreement with the assignments in Table 4. In contrast, in the Z form of CG8Ph, the phenyl group lies outside of the helix and the phenyl-H2/6 equivalent and phenylH3/5 protons are equivalent. Thus, in this case, only three different resonances are expected, in agreement with the data for the Z DNA conformer. The models provide the means to determine the average distances between protons on the phenyl ring and those on C5, G6, and C7. The shorter distances (less than 3 Å) determined from the most representative structure for the B DNA between the phenyl protons and the rest of the oligonucleotide are phenyl-H2/G6-H2′ (2.22 Å), phenyl-H2/G6-H3′ (2.64 Å), phenyl-H4/G6-H2′ (2.68 Å), phenyl-H4/C5-H2 ′′ (2.70 Å), phenyl-H5/C5-H6 (2.72 Å), and phenyl-H5/C5-H2′ (2.74 Å). Of these pairs, we see NOESY correlations from phenyl-H2/G6-H2′ (2.22 Å) and phenyl-H2/G6-H3′ (2.64 Å), i.e., those with the shortest calculated differences. Unfortunately, the only NOESY correlations that we clearly observe were to the modified bases’ own sugar protons. We did not observe any crosspeaks between the phenyl protons and any of the sugar protons of C5, which would help to establish firmly the conformation of the glycosidic bond of G6. However, the anti conformation seems most likely, since the base pairing in the C:G8Ph base pair would be lost and would significantly destabilize the duplex. In addition, this conformation is consistent with the large upfield shifts of the phenyl-H2 and H3 protons. The most representative structure for the Z DNA form of CG8Ph is shown in Figure 5b. In the Z DNA form, the phenyl group is completely outside of the helix and it can freely rotate, rendering the ortho and the meta protons equivalent (Table 4). We calculated the RMSD distances between the phenyl protons and those in the oligonucleotide. Only one distance is less than 3 Å, Ph-H2/ C5-H1′ (2.85 Å). Two additional distances of interest are Ph-H6/G6-H2′ (3.31 Å) and Ph-H6/C5-H1′ (3.30 Å). Of these pairs, only the Ph-H2/C5-H1′ and Ph-H6/C5-H1′ were observed.

Discussion In this work, we have begun to develop a general synthetic approach to the preparation of C8-arylguanine derivatives. These derivatives can be converted to the corresponding phosphoramidite and used for the preparation of oligonucleotides. The previous synthetic route to C8-arylguanines required the base-induced decomposition of arenediazonium ions in the presence of 2′deoxyguanosine (28, 32). However, this reaction is difficult to conduct reproducibly, the products require extensive purification, and the overall yield is poor. In addition, the C8-aryl-2′-deoxyadenosines cannot be prepared by this route, as deribosylation occurs and only the C8aryladenine derivatives are obtained (28). The new

1392 Chem. Res. Toxicol., Vol. 16, No. 10, 2003

method described here is based on the aqueous version of the Suzuki coupling reaction (52). This route appears to be quite general for the preparation of C8-aryl-2′deoxyguanosines as well as C8-aryl-2′-deoxyadenosines (53). The products are produced in fair to good yields and are easily purified. Moreover, this method should be generally applicable and other C8-arylpurines, such as those adducts formed from polyaromatic hydrocarbons (16, 54) should be accessible via this route. Thus, this methodology should facilitate studies involving C8-aryl adducts such as the effect of C8-arylguanine substitution on B/Z DNA equilibrium. There are a number of factors known to affect the relative stability of B and Z DNA including salt concentration, identity of the cation, sequence effects, base modifications, and hydrophobicity. All of these effects are important and, to some extent, overlap with one another. Here, the effects that most likely need to be considered are base modifications, salt concentration, and hydrophobicity. C8-guanine modification is known to stabilize the syn glycosidic bond conformation due to an unfavorable steric interaction between groups in the C8-guanine position and the H-2′′ proton (55). This corresponds to the conformation of purines in Z DNA; therefore, C8phenylguanosines will stabilize the Z DNA conformation and destabilize the B DNA conformation where the preferred glycosidic bond conformation for all unmodified bases is anti. A second steric factor that may be important and that may destabilize the B DNA form results from the fact that the phenyl group of CG8Ph lies in the major groove and lies outside of the helix in the Z form. Finally, there are differences in the hydrophobic interactions of the phenyl group in the B and Z DNA forms. In the B form, the phenyl group is largely shielded from the solvent (water) and fully exposed to solvent in the Z form. Unlike steric effects, here, hydrophobic effects will tend to stabilize the B form relative to the Z form. Only one example of a C8-phenylguanine modified oligonucleotide has been studied by thermal denaturation (32). In that case, the C8-phenylguanine modification was found to decrease the Tm. In the present case, we do not see any difference in the Tm between CG and CG8Ph. Under low salt conditions, CG8Ph is not in one form; rather, it is a mixture of three forms, although, as the melting temperature is approached, the B form predominates. Consequently, the reported melting temperature may mainly reflect the melting of the B form. Nevertheless, significant ambiguities exist and whether the C8phenylguanine modification destabilizes oligonucleotides in the B DNA form will require additional studies. Under high salt conditions, where the Z DNA form of both CG and CG8Ph is present, the melting temperatures measured were identical and suggest that the Z DNA forms of CG and CG8Ph are of similar stability. This is not unexpected as in the Z DNA conformation, the phenyl group of the modified oligonucleotide is pointed away from the duplex, and the phenyl group cannot sterically destabilize the duplex. Although the thermal denaturation data cannot be used to quantitate the effect of the phenyl modification on the B DNA form, the C8-phenylguanine modification does stabilize the Z DNA form CG8Ph relative to the Z DNA form of CG. This is clearly demonstrated by the CD data that show that a 1:1 mixture of the B and Z DNA forms in CG requires a salt concentration 5-10 times higher than required for CG8Ph. This result is comparable

Gannett et al.

to what has been observed in recent studies with alternating purine/pyrimidine oligonucleotides with C8-bromo (56) or C8-methyl (33) substitution although these latter two modifications appear to be even more effective at shifting the B/Z equilibrium in favor of the Z DNA form. The difference between the bromo or methyl substituents and the phenyl substituent may be due to the spherical nature of the former and the planer nature of the latter. In addition, because the bromo and methyl substituents have less surface area than a phenyl group, unfavorable hydrophobic interactions may be less for the bromo and methyl substituents than for the phenyl group. Nevertheless, it is interesting that the Z DNA form is adopted under physiological conditions, as the steric effects of the C8-guanine substitution by bromine, methyl, or phenyl must overcome the unfavorable electrostatic interactions present in the Z form. Bromo, methyl, and phenyl modifications, therefore, may be useful in the study of some of the factors that affect the B/Z equilibrium. The NMR data further support the CD data regarding the B/Z equilibrium. The NMR data acquired at 1 M sodium chloride confirm the complete conversion to the Z form. Furthermore, at lower salt concentrations (25200 mM), the CD data suggest and the NMR data unequivocally show that CG8Ph exists as a mixture of B and Z. A third species is also present that has been tentatively assigned as the single-stranded form. At least two additional possibilities exist for this third form. First, Z DNA is known to exist in different forms (ZI/ZII (57) and Z[WC] (58)). It is not clear that the phenyl groups in these different forms would have different chemical shifts, but if so, then two different Z DNA forms could be present. However, the assignment of the three forms to single stranded, B DNA, and Z DNA seems the most likely possibility based on the variable temperature NMR data. Previous studies have shown that the Z DNA form is preferred at low temperature. Upon raising the temperature, the concentration of the B DNA form increases at the expense of the Z form. The single-stranded form also increases in concentration with temperature, albeit more slowly than the B DNA form, until the Tm is reached. The variable temperature NMR data for CG8Ph are in agreement with this. The presence of all three forms complicates assignment of the NMR data. In addition, because there are multiple forms of the modified oligonucleotide at low salt concentrations, the thermal denaturation data cannot be used to determine thermodynamic parameters and then compared to the corresponding unmodified oligonucleotide. We are currently exploring several alternatives that may reduce the complexity of the system. First, oligonucleotides are being prepared with the C8-phenylguanine modified base located at G8/G18. This may reduce any unfavorable interactions caused by having the modified base pairs adjacent to one another. A second approach is to use a sequence that is not self-complementary so that the duplex will only contain one C8phenylguanine but that is still prone to Z DNA formation. A sequence meeting these criteria is 3′TGTGTG8PhTGTG5′. The resulting duplex will contain only one G8Ph modified base, and the duplex with 3′CACACACACA5′ has been studied and is known to be prone to adopting a Z DNA conformation, albeit less prone than CG (59-62). A final approach that may be explored is to stay with the same sequence and prepare oligonucleotides with other C8arylguanine modifications including p-tolyl, p-hydroxy-

C8-Phenylguanine Modified Oligonucleotides

methylphenyl, and p-carboxyphenyl, corresponding to the adducts that have been observed from the corresponding hydrazines in vitro and in vivo. Because of similar steric properties but significantly different hydrophobicities, this oligos may show different B/Z preferences. Because of the complexity of the NMR data at low salt concentration, a complete molecular model cannot, at this time, be derived from NOE-based distance constraints. However, the limited structural information that we have for CG8Ph, derived from the NMR data, are consistent with the structural models shown in Figure 5. In particular, the chemical shifts of the phenyl protons ascribed to the B and Z forms are consistent with the molecular models and the latter help to understand the observed chemical shifts. In addition, the data suggest that the B form adopts an anti conformation about the modified base’s glycosidic bond. This conformation places the phenyl group in the major groove, and all five phenyl protons are in different environments leading to the prediction that all five will have different chemical shifts. Finally, the alternative syn conformation of the phenyl modified G (G6) in the B DNA conformer cannot be rigorously excluded. However, because the loss of the C:G8Ph base pair would significantly destabilize the duplex, this seems unlikely. Finally, the finding that C8-phenylguanine modified purine/pyrimidine tracts are prone to Z DNA formation may be related to the carcinogenic nature of aryl hydrazines. All of the aryl hydrazines that we have studied have been found to produce C8-aryl purine adducts. Studies of C8-phenylguanine adducts suggest that misreading is inefficient, so this type of mechanism for carcinogenesis seems unlikely. An alternative model is stabilization of the Z DNA form. A recent study by Liu et al. (34) demonstrated that activation of a gene (CSF1) promoter requires adoption of a Z DNA conformation. While unmodified Z DNA prone sequences (alternating pyrimidine/purine sequences) adopt a B DNA conformation under physiological conditions, C8-arylation of guanine may alter the preference and stabilize the Z DNA structure required for gene expression. Stabilization of the Z form would be further enhanced if the G8Ph adduct formed in a sequence containing C5Me. C5Me is known to stabilize Z DNA in CG sequences (43), and the level of C5Me is correlated with gene expression. Because Z DNA prone sequences have been identified in oncogenes such as c-Myc, it is possible that the C8-arylguanine adducts formed from carcinogenic aryl hydrazines activate these genes by this mechanism.

Acknowledgment. We thank the National Science Foundation (NSF 1002165R) for their support of this work, the Florida State University Bioanalytical Analysis, Synthesis, and Sequencing (BASS) Laboratory for the preparation of the oligonucleotides used in this work, and both the Florida State University NMR Laboratory and the National High Field Magnetic Laboratory for providing access to their NMR instrumentation and their assistance with the NMR studies. Supporting Information Available: Development of force field parameters and force field parameters. This material is available free of charge via the Internet at http://pubs.acs.org.

References (1) Heflich, R. H., and Neft, R. E. (1994) Mutat. Res. Rev. Gen. Toxicol. 318, 73-174.

Chem. Res. Toxicol., Vol. 16, No. 10, 2003 1393 (2) Le Page, F., Margot, A., Grollman, A. P., Sarasin, A., and Gentil, A. (1995) Mutagenicity of a Unique 8-Oxoguanine in a Human Ha-ras Sequence in Mammalian Cells. Carcinogenesis 16, 27792784. (3) Poltev, V. I., Smirnov, S. L., Issarafutdinova, O. V., and Lavery, R. (1993) Conformations of DNA Duplexes Containing 8-Oxoguanine. J. Biomol. Struct. Dyn. 11, 293-301. (4) Yermilov, V., Rubio, J., Becchi, M., Friesen, M. D., Pignatelli, B., and Ohshima, H. (1995) Formation of 8-Nitroguanine by the Reaction of Guanine with Peroxynitrite in vitro. Carcinogenesis 16, 2045-2050. (5) Shafirovich, V., Mock, S., Kolbanovskiy, A., and Geacintov, N. E. (2002) Photochemically Catalyzed Generation of Site-Specific 8-Nitroguanine Adducts in DNA by the Reaction of Long-lived Neutral Guanine Radicals with Nitrogen Dioxide. Chem. Res. Toxicol. 15, 591-597. (6) Kouchakdjian, M., Bodepudi, V., Shibutani, S., Eisenberg, M., Johnson, F., Grollman, A. P., and Patel, D. J. (1991) NMR Structural Studies of the Ionizing Radiation Adduct 7-Hydro8-oxodeoxyguanosine (8-oxo-7H-dG) Opposite Deoxyadenosine in a DNA Duplex. 8-Oxo-7H-dG(syn).dA(anti) Alignment at Lesion Site. Biochemistry 30, 1403-1412. (7) Shibutani, S., Takeshita, M., and Grollman, A. P. (1991) Insertion of Specific Bases During DNA Synthesis Past the Oxidationdamaged Base 8-OxodG. Nature 349, 431-433. (8) Retel, J., Hoebee, B., Braun, J. E. F., Lutgerink, J. T., van den Akker, E., Wanamarta, A. H., Joenje, H., and Lafleur, M. V. M. (1993) Mutational Specificity of Oxidative DNA Damage. Mutat. Res. 299, 165-182. (9) Thiviyanathan, V., Somasunderam, A., Hazra, T. K., Mitra, S., and Gorenstein, D. G. (2003) Solution Structure of a DNA Duplex Containing 8-Hydroxy-2′-Deoxyguanosine Opposite Deoxyguanosine. J. Mol. Biol. 325, 433-442. (10) O’Handley, S. F., Sanford, D. G., Xu, R., Lester, C. C., Hingerty, B. E., Broyde, S., and Krugh, T. R. (1993) Structural Characterization of an N-Acetyl-2-aminofluorene (AAF) Modified DNA Oligomer by NMR, Energy Minimization, and Molecular Dynamics. Biochemistry 32, 2481-2497. (11) Augusto, O., Cavalieri, E. L., Rogan, E. G., RamaKrishna, N. V. S., and Kolar, C. (1990) Formation of 8-Methylguanine as a Result of DNA Alkylation by Methyl Radicals Generated during Horseradish Peroxidase-catalyzed Oxidation of Methylhydrazine. J. Biol. Chem. 265, 22093-22096. (12) Gannett, P. M., Lawson, T., Miller, M., Thakkar, D. D., Lord, J. W., Yau, W.-M., and Toth, B. (1996) 8-Arylguanine Adducts from Arenediazonium Ions and DNA. Chem-Biol. Interact. 95, 1-25. (13) Gannett, P. M., Powell, J. H., Rao, R., Shi, X., Lawson, T., Kolar, C., and Toth, B. (1999) C8-Arylguanine and C8-Aryladenine Formation in Calf Thymus DNA from Arenediazonium Ions. Chem. Res. Toxicol. 12, 297-304. (14) Powell, J. H., and Gannett, P. M. (2002) Mechanisms of Carcinogenicity of Aryl Hydrazines, Aryl Hydrazides, and Arenediazonium Ions. J. Exp. Pathol. Toxicol. Oncol. 21, 1-31. (15) Gannett, P. M., Yau, W.-M., Lawson, T., Lord, J., Kolar, C., and Toth, B. (1997) Formation of C8-Aryladenine Adducts from Arenediazonium ions and DNA. Proc. Am. Assoc. Cancer Res. 38, A2233. (16) RamaKrishna, N. V. S., Gao, F., Padmavathi, N. S., Cavalieri, E., Rogan, E. G., Cerny, R. L., and Gross, M. L. (1992) Model Adducts of Benzo[a]pyrene and Nucleosides Formed from Its Radical Cation and Diol Epoxide. Chem. Res. Toxicol. 5, 293302. (17) Kikugawa, K., Kato, T., and Kojima, K. (1992) Substitution of pand o-Hydroxyphenyl Radicals at the 8 Position of Purine Nucleosides by Reaction with Mutagenic p- and o-Diazoquinones. Mutat. Res. 268, 65-75. (18) Nakao, L. S., and Augusto, O. (1998) Nucleic Acid Alkylation by Free Radical Metabolites of Ethanol. Formation of 8-(1-Hydroxyethyl)guanine and 8-(2-Hydroxyethyl)guanine Adducts. Chem. Res. Toxicol. 11, 888-894. (19) Nakao, L. S., Fonseca, E., and Augusto, O. (2002) Detection of C8-(1-Hydroxyethyl)guanine in Liver RNA and DNA from Control and Ethanol-Treated Rats. Chem. Res. Toxicol. 15, 1248-1253. (20) Hix, S., Kadiiska, M. B., Mason, R. P., and Augusto, O. (2000) In Vivo Metabolism of tert-Butyl Hydroperoxide to Methyl Radicals. EPR Spin-Trapping and DNA Methylation Studies. Chem. Res. Toxicol. 13, 1056-1064. (21) Stiborava, M., Asfaw, B., Anzenbacher, P., Leseticky, L., and Hodek, P. (1988) The first identification of the benzenediazonium ion formation from a non-aminoazo Dye, 1-Phenylazo-2-hydroxynaphthalene (Sudan I) by microsomes of rat livers. Cancer Lett. 40, 319-326.

1394 Chem. Res. Toxicol., Vol. 16, No. 10, 2003 (22) Stiborava, M., Asfaw, B., Anzenbacher, P., and Hodek, P. (1988) A New Way to Carcinogenicity of Azo Dyes: The Benzenediazonium ion formed from a Non-aminoazo Dye, 1- Phenylazo-2hydroxynaphthalene (Sudan I) by Microsomal Enzymes Binds to Deoxyguanosine Residues of DNA. Cancer Lett. 40, 327-333. (23) Malaveille, C., Brun, G., Kolar, G., and Bartsch, H. (1982) Mutagenic and Alkylating Activities of 3-Methyl-1- phenyltriazenes and their possible role as carcingenic metabolites of the parent dimethyl compounds. Cancer Res. 42, 1446-1453. (24) Freese, E., Sklarow, S., and Freese, E. B. (1968) DNA Damage Caused by Antidepressant Hydrazines and Related Drugs. Mutat. Res. 5, 343-348. (25) Williams, G. M., Mazue, G., and McQueen, C. A. (1980) Genotoxicity of the Antihypertensive Drugs Hydralazine and Dihydralazine. Science 210, 329-330. (26) Dokka, S., and Rojanasakul, Y. (2000) Novel Nonendocytic Delivery of Antisense Oligonucleotides. Adv. Drug Delivery Rev. 44, 35-49. (27) Toth, B. (2000) A review of the natural occurrence, synthetic production and use of carcinogenic hydrazines and related Chemicals. In Vivo 14, 299-320. (28) Lawson, T., Gannett, P. M., Yau, W.-M., Dalal, N. S., and Toth, B. (1995) Different Patterns of Mutagenicity of Arenediazonium Ions in V79 Cells and Salmonella Typhimurium TA 102: Evidence for Different Mechanisms of Action. J. Agric. Food Chem. 43, 2627-2635. (29) Lawson, T. (1987) Metabolism of Arylhydrazines by Cytochrome P-450 Mixed Function Oxidases and Prostaglandin(H)synthase from Mouse Lungs. Cancer Lett. 34, 193-200. (30) Lawson, T., and Chauhan, Y. (1985) Metabolism of Arylhydrazines by Mouse Liver Mixed-Function Oxidases In Vitro. J. Agric. Food Chem. 33, 218-219. (31) Gannett, P. M., Shi, X., Lawson, T., Kolar, C., and Toth, B. (1997) Aryl Radical Formation During the Metabolism of Arylhydrazines by Microsomes. Chem. Res. Toxicol. 10, 1372-1377. (32) Kohda, K., Tsunomoto, H., Kasamatsu, T., Sawamura, F., Terashima, I., and Shibutani, S. (1997) Synthesis and Miscoding Specificity of Oligodeoxynucleotide Containing 8-Phenyl-2′-deoxyguanosine. Chem. Res. Toxicol. 10, 1351-1358. (33) Sugiyama, H., Kawai, K., Matsunaga, A., Fujimoto, K., Saito, I., Robinson, H., and Wang, A. H. J. (1996) Synthesis, Structure and Thermodynamic Properties of 8-Methylguanine-containing Oligonucleotides: Z-DNA Under Physiological Salt Conditions. Nucleic Acids Res. 24, 1272-1278. (34) Liu, R., Liu, H., Chen, X., Kirby, M., Brown, P. O., and Zhao, K. (2001) Regulation of CSF1 Promoter by the SWI/SNF-like BAF Complex. Cell 106, 309-318. (35) Gannett, P. M., and Sura, T. P. (1993) An Improved Synthesis of 8-Bromo-2′-Deoxyguanosine. Synth. Commun. 23, 1611-1615. (36) Cornell, W. D., Cieplak, P., Bayly, C. I., Gould, I. R., Merz, K. M., Jr., Ferguson, D. M., Spellmeyer, D. C., Fox, T., Caldwell, J. W., and Kollman, P. A. (1999) A Second Generation Force Field for the Simulation of Proteins, Nucleic Acids, and Organic Molecules. J. Am. Chem. Soc. 117, 5179-5197. (37) Simmerling, C., Elber, R., and Zhang, J. (1995) MOIL-ViewsA Program for Visualization of Structure and Dynamics of Biomoleculessand STOsA Program for Computing Stochastic Paths. In Modeling of Bimolecular Structure and Mechanisms (Pullman, A., Ed.) pp 241-265, Kluwer, The Netherlands. (38) Miyaura, N., and Suzuki, A. (1995) Palladium-catalyzed Crosscoupling Reactions of Organoboron Compounds. Chem. Rev. 95, 2457-2483. (39) Amann, N., and Wagenknecht, H.-A. (2002) Preparation of Pyrenyl-modified Nucleosides via Suzuki-Miyaura Cross-coupling Reactions. Synlett 687-691. (40) Lakshman, M. K., Thomson, P. F., Nuqui, M. A., Sevova, N., and Boggess, B. (2002) Facile Pd-catalyzed Cross-coupling of 2′Deoxyguanosine O6-Arylsulfonates with Arylboronic Acids. Org. Lett. 4, 1479-1482. (41) Dupuis, C., Adiey, K., Charruault, L., Michelet, V., Savignac, M., and Genet, J. P. (2001) Suzuki Cross-coupling of Arylboronic Acids Mediated by a Hydrosoluble Pd(0)/TPPTS Catalyst. Tetrahedron Lett. 42, 6523-6526. (42) Casalnuovo, A. L., and Calabrese, J. C. (1990) PalladiumCatalyzed Alkylations in Aqueous Media. J. Am. Chem. Soc. 112, 4324-4330.

Gannett et al. (43) Behe, M., and Felsenfeld, G. (1981) Effects of Methylation on a Synthetic Polynucleotide: The B-Z Transition in Poly(dG-m5dC)‚ poly(dG-m5dC). Biochemistry 23, 54-62. (44) Johnson, W. C. (2000) CD of Nucleic Acids. In Circular Dichroism: Principles and Applications (Berova, N., Nakanishi, K., and Woody, R. W., Eds.) pp 703-718, John Wiley and Sons, NY. (45) Abuaf, P., Kadlubar, F. F., and Grunberger, D. (1987) Circular Dichroism of Poly(dG-dC) Modified by the Carcinogens N-Methyl4-Aminoazobenzene or 4-Aminobiphenyl. Nucleic Acids Res. 15, 7125-7136. (46) van Houte, L. P. A., Westra, J. G., and van Grondelle, R. A. (1988) Spectroscopic Study of the Conformation of Poly d(G-C)‚poly d(G-C) Modified with the Carcinogen 2-Aminofluorene. Carcingenesis 9, 1017-1027. (47) Fazakerley, G. V., van der Marel, G. A., van Boom, J. H., and Guschlbauer, W. (1984) Helix Opening in Deoxyribonucleic Acid from a Proton Nuclear Magnetic Resonance Study of Imino and Amino Protons in d(CG)3. Nucleic Acids Res. 12, 8269-8279. (48) Orbons, L. P. M., van der Marel, G. A., van Boom, J. H., and Altona, C. (1986) The B and Z Forms of the d(m5C-G)3 and d(br5C-G)3 Hexamers in Solution. A 300-MHz and 500-MHz TwoDimensional NMR Study. Eur. J. Biochem. 160, 131-139. (49) Rocchia, W., Alexov, E., and Honig, B. (2001) Extending the Applicability of the Nonlinear Poisson-Boltzmann Equation: Multiple Dielectric Constants and Multivalent Ions. J. Phys. Chem. B 105, 6507-6514. (50) Cheatham, I. T. E., Srinivasan, J., Case, D. A. M., and Kollman, P. A. (1998) Molecular Dynamics and Continuum Solvent Studies of the Stability of PolyG-PolyC and PolyA-PolyT DNA Duplexes in Solution. J. Biomol. Struct. Dyn. 16, 671-682. (51) Tsui, V., and Case, D. A. (2000) Molecular Dynamics Simulations of Nucleic Acids with a Generalized Born Solvation Model. J. Am. Chem. Soc. 122, 2489-2498. (52) Shaughnessy, K. H., and Booth, R. S. (2001) Sterically Demanding, Water-Soluble Alkylphosphines as Ligands for High Activity Suzuki Coupling of Aryl Bromides in Aqueous Solvents. Org. Lett. 3, 2757-2759. (53) Western, E. C., Daft, J. R., Johnson, E. M., II, Gannett, P. M., and Shaughnessy, K. H. (2003) Efficient One-Step Suzuki Arylation of Unprotected Halonucleosides using Water-Soluble Palladium Catalysts. J. Org. Chem. 68, 6767-6774. (54) Chakrabarti, M. C., and Schwarz, F. P. (1999) Thermal Stability of PNA/DNA and DNA/DNA Duplexes by Differential Scanning Calorimetry. Nucleic Acids Res. 27, 4801-4806. (55) Stolarski, R., Dudycz, L., and Shugar, D. (1980) NMR Studies on the syn-anti Dynamic Equilibrium in Purine Nucleosides and Nucleotides. Eur. J. Biochem. 108, 111-121. (56) Fabrega, C., Macias, M. J., and Eritja, R. (2001) Synthesis and Properties of Oligonucleotides Containing 8-Bromo-2′-deoxyguanosine. Nucleosides, Nucleotides, Nucleic Acids 20, 251-260. (57) Basham, B., Eichman, B. F., and Ho, P. S. (1999) The Singlecrystal Structures of Z-DNA. In Oxford Handbook of Nucleic Acid Structure (Neidle, S., Ed.) pp 199-252, Oxford University Press, Oxford. (58) Ansevin, A. T., and Wang, A. H. (1990) Evidence for a new Z-type Left-handed DNA Helix: Properties of Z(WC)-DNA. Nucleic Acids Res. 18, 6119-6126. (59) Zimmer, C., Tymen, S., Marck, C., and Guschlbauer, W. (1982) Conformational Transitions of Poly(dA-dC)‚poly(dG-dT) Induced by High Salt or in Ethanolic Solution. Nucleic Acids Res. 10, 1081-1091. (60) Thomas, T. J., and Messner, R. P. (1986) A Left-handed (Z) Conformation of Poly(dA-dC)‚poly(dG-dT) Induced by Polyamines. Nucleic Acids Res. 14, 6721-6733. (61) Riazance-Lawrence, J. H., and Johnson, W. C., Jr. (1992) Multivalent Ions Are Necessary for Poly[d(AC).d(GT)] to Assume the Z form: A CD Study. Biopolymers 32, 271-276. (62) Reich, Z., Friedman, P., Levin-Zaidman, S., and Minsky, A. (1993) Effects of Adenine Tracts on the B-Z Transition. J. Biol. Chem. 268, 8261-8266.

TX034023D