Nucleosides and Nucleotides. 218. Alternate-Strand Triple-Helix

shift assay (EMSA). It was found that the alternate-strand triplexes were significantly stabilized by linking the TFO fragments with the pentaerythrit...
0 downloads 0 Views 154KB Size
Download PDF
Bioconjugate Chem. 2003, 14, 607−613

607

Nucleosides and Nucleotides. 218. Alternate-Strand Triple-Helix Formation by the 3′-3′-Linked Oligodeoxynucleotides Using a Purine Motif† Shuichi Hoshika, Yoshihito Ueno,‡ and Akira Matsuda* Graduate School of Pharmaceutical Sciences, Hokkaido University, Kita-12, Nishi-6, Kita-ku, Sapporo 060-0812, Japan. Received December 10, 2002

In this paper, we describe the synthesis of the 3′-3′-linked TFOs that can form the antiparallel triplexes with the duplex DNA target by reverse Hoogsteen hydrogen bonds. Stability of the alternate-strand triplexes between these TFOs and the target DNAs was investigated using the electrophoretic mobility shift assay (EMSA). It was found that the alternate-strand triplexes were significantly stabilized by linking the TFO fragments with the pentaerythritol linker. And, unlike the alternate-strand triplexes composed of the pyrimidine motif, the terminal ammonium ion of the aminobutyl-linker and the intercalator of the TFOs did not contribute to the stability of the alternate-strand triplex comprised of the purine motif. We also tested the ability of the 3′-3′-linked TFOs to inhibit cleavage of the duplex DNA target 17 by the restriction enzyme EcoT14I and found that the 3′-3′-linked TFOs 12 and 13 inhibited the cleavage by the enzyme more effectively than the unlinked decamer 8. Thus, the TFOs linked with pentaerythritol may be useful as the antigene oligonucleotide to the DNA targets, which have alternating oligopyrimidine-oligopurine sequences.

INTRODUCTION

Triple helix (triplex)-forming oligonucleotides (TFOs) have attracted a great deal of attention because of their ability to specifically bind double-stranded DNA and because of their potential use in gene therapy (i.e., in antigene strategy). A triplex is usually formed through sequence-specific interaction of a single-stranded oligopurine or oligopyrimidine TFO with a major groove of oligopurine/oligopyrimidine stretch in DNA duplex (1). In the pyrimidine motif triplex, an oligopyrimidine TFO binds parallel to the oligopurine strand of the target duplex by Hoogsteen hydrogen bonding to form T‚AT and C+‚GC base triplets (2, 3). On the other hand, in the purine motif triplex, an oligopurine TFO binds antiparallel to the oligopurine strand of the target by reverse Hoogsteen hydrogen bonding to form A‚AT (or T‚AT) and G‚GC base triplets (4, 5). However, target sequences in the antigene strategy are very restricted. Since thermal stability of the triplexes is generally much lower than that of the duplexes under physiological conditions, an oligopurine cluster with long chain lengths is usually required for stable triplex formation. In addition, formation of the pyrimidine motif triplex needs conditions of low pH (pH < 6.0) because unmodified cytosine residues, if present in the TFO, must be protonated to bind with guanine of the G:C duplex (2, 6). Several approaches have been attempted to expand the repertory of potential DNA targets. If short oligopurine sequences appear adjacently and alternately on the two strands of the DNA target, the alternate sequences can † For Part 217 in this series, see: Minakawa, N., Kato, Y., Uetake, K., Kaga, D., and Matsuda, A. (2003) An improved large scale synthesis of 1,4-anhydro-4-thio-D-ribitol. Tetrahedron 59, 1699-1702. * Corresponding author. Phone: +81-11-706-3228. Fax: +8111-706-4980. E-mail: [email protected]. ‡ Present address: Faculty of Engineering, Gifu University, Yanagido, Gifu 501-1193, Japan.

be recognized by TFOs that simultaneously and cooperatively bind to the adjacent oligopurine domains on alternate strands by crossover in the major groove (724). These so-called “alternate-strand triplex formations” are divided into two categories. A natural oligonucleotide such as 5′-oligopurine-oligopyrimidine-3′ or 5′-oligopyrimidine-oligopurine-3′ can be used as a TFO for alternatestrand triplex formation (7-14). In this case, the oligopyrimidine part of the TFO binds to the oligopurine domain on the DNA target by Hoogsteen hydrogen bonding, and the oligopurine part binds to the oligopurine domain by reverse Hoogsteen hydrogen bonding. In the second approach, only one set of motifs (i.e., Hoogsteen or reverse Hoogsteen motif) is employed (15-24). In this case, the 3′- and 3′- or 5′- and 5′-termini of the two oligonucleotide fragments have to be connected by using an appropriate linker to invert polarity. The different possibilities for alternate-strand triplex formations in the latter case are shown in Figure 1. When the oligopyrimidine sequence follows the oligopurine sequence, the 3′-3′-linked and the 5′-5′-linked TFOs can form the parallel (a) and antiparallel (b) triplexes with the duplex DNA target by Hoogsteen and reverse Hoogsteen hydrogen bonds, respectively. On the other hand, when the oligopurine sequence follows the oligopyrimidine sequence, the 5′-5′-linked and the 3′3′-linked TFOs can form the parallel (c) and antiparallel (d) triplexes with the duplex DNA target by Hoogsteen and reverse Hoogsteen hydrogen bonds, respectively. Recently, we reported the synthesis of the 3′-3′-linked TFOs with the 2-anthraquinonecarbonyl group at the junction point (25). We found that the 3′-3′-linked TFOs with the 2-anthraquinonecarbonyl group as an intercalator enhanced the thermal stability of the parallel type of triplex with a DNA target (Figure 1a) as compared with those without the intercalator and the unlinked nonamer. Moreover, we revealed that the 3′-3′-linked TFO with the anthraquinonyl group inhibits the cleavage by the restriction enzyme Hind III more effectively than the unlinked nonamer and the 3′-3′-linked TFO without

10.1021/bc0256493 CCC: $25.00 © 2003 American Chemical Society Published on Web 04/02/2003

608 Bioconjugate Chem., Vol. 14, No. 3, 2003

Hoshika et al.

Figure 1. Schematic presentation of “alternate-strand triplex” formation by the 3′-3′- or 5′-5′-linked TFOs. Scheme 1

a

a (a) DMTrCl, pyridine, room temperature; (b) (1) Bu SnO, benzene, 80 °C; (2) BzCl, benzene, room temperature; (c) succinic 2 anhydride, Et3N, DMAP, CH3CN, room temperature; (f) aminopropyl-controlled pore glass, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride, DMF, room temperature.

the intercalator. However, as mentioned above, because cytosines have to be protonated, the pyrimidine-rich motif requires a slightly acidic pH for stable triplex formation, thereby limiting its use under physiological conditions. In contrast, triplex formation by purine-rich TFOs is pHindependent, and for this reason they have been more frequently used in the antigene strategy (1, 26, 27). In this paper, we report the synthesis of the 3′-3′-linked TFOs that can form the antiparallel triplexes with the duplex DNA target by reverse Hoogsteen hydrogen bonds (Figure 1d). Stability of the alternate-strand triplexes between these TFOs and the target DNAs was investigated with the use of the electrophoretic mobility shift assay (EMSA). The ability of the 3′-3′-linked TFOs to inhibit cleavage of a target DNA by a restriction enzyme was also studied. RESULTS AND DISCUSSION

Synthesis. The controlled pore glass (CPG) with Pentaerythritol was synthesized as shown in Scheme 1.

First, pentaerythritol 1 was treated with two equivalents of 4,4′-dimethoxytrityl chloride (DMTrCl) in pyridine to give the O-bis(DMTr) derivative 2 in 27% yield. Compound 2 was treated with Bu2SnO and then reacted with BzCl to produce the O-mono(Bz) derivative 3 in 98% yield. Compound 3 was succinated to give 4, which was further reacted with CPG to afford a solid support 5 bearing 4 (25 µmol/g). The 3′-3′-linked TFOs 13-16 (Figure 2a), which consist of 2′-deoxyguanosine (G) and thymidine (T) or G and 2′deoxyadenosine (A), were synthesized on a DNA synthesizer using the CPG 5. The fully protected TFOs (each 0.25 µmol) linked to the solid support were treated with concentrated NH4OH at 55 °C for 16 h. The released TFOs were purified by denaturing with 20% polyacrylamide gel electrophoresis (20% PAGE) to give the deprotected TFOs 13-16, in 6.6, 9.9, 6.7, and 7.1 OD260 units, respectively. These TFOs were analyzed by matrixassisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF/MS), and the observed mo-

Alternate-Strand Triple Helix

Bioconjugate Chem., Vol. 14, No. 3, 2003 609

Figure 2. Sequences of TFOs and the target duplexes.

Figure 3. Quantitative EMSA to detect binding of TFO 8 (lanes 3-8), TFO 10 (lanes 9-16), TFO 11 (lanes 17-24), and TFO 12 (lanes 25-32) to DNA target. DNA concentration: 10 nM. TFO concentrations are indicated.

lecular weights supported their structures. The TFO 10 with an aminobutyl-linker, the TFO 11 protected with an acetyl group, and the TFO 12 with a 2-anthraquinonecarbonyl group were also synthesized according to the reported method (25) using the CPGs 6 and 7 (Scheme 1). Studies of Triplex Formation by the Electrophoretic Mobility Shift Assay. The stability of the triplexes was investigated using the electrophoretic mobility shift assay (EMSA). The EMSA detects the difference in electrophoretic mobility of the doublestranded DNA and the triplex formed. First, we tested the ability of the unlinked decamers 8 and 9 and the 3′3′-linked TFOs 10-12 to bind to the target DNA 17, which has oligopurine sequences, one on each strand (Figure 2b). The sequences of the binding domains (each 10 mer length) are symmetrical. The 32P-labeled DNA 17 was incubated in the presence of increasing concentrations of the unlinked decamer 8 or 9 or the 3′-3′-linked TFO 10, 11, or 12 in a buffer of 20 mM Tris-HCl (pH 7.5) containing 10 mM MgCl2. The mixtures were then analyzed by nondenaturing 20% polyacrylamide gel electrophoresis (PAGE) at 4 °C. As shown in Figure 3, the retarded bands corresponding to the triplexes were observed for the TFOs 10, 11, and 12, whereas the band corresponding to the triplex was not detected for the TFOs 8 and 9. Table 1 summarizes the dissociation constants (Kds) of the TFOs bound to the duplex DNA target. The Kd values of the TFOs 10, 11,

Table 1. Summary of Dissociation Constants (Kd) of the TFOs Bound to the DNA Targets TFO

duplex

Kd (nM)

TFO 8 TFO 9 TFO 10 TFO 11 TFO 12 TFO 13 TFO 14 TFO 15 TFO 16 TFO 13

DNA 17 DNA 17 DNA 17 DNA 17 DNA 17 DNA 17 DNA 17 DNA 17 DNA 17 DNA 18

>1 000 000 >1 000 000 33 ((12) 33 ((8) 38 ((22) 41 ((6) 17 000 ((3000) 340 ((10) >100 000 >100 000

and 12 to the DNA 17 were 33 ( 12, 33 ( 8, and 38 ( 22 nM, respectively. On the other hand, the Kd values of the unlinked decamers 8 and 9 to the DNA 17 were both more than 1 mM. It was found that the alternate-strand triplexes are significantly stabilized by linking the TFO fragments with the pentaerythritol linker. However, the Kd values of TFO 10 with the aminobutyl-linker, 11 with N-acetylaminobutyl-linker, and 12 with the 2-anthraquinonecarbonyl group to the DNA 17 were almost the same. Thus, in contrast to the alternate-strand triplexes composed of the pyrimidine motif (25), the terminal ammonium ion of the aminobutyl-linker and the intercalator of the TFOs were found not to contribute to the stability of the alternate-strand triplex consisting of the purine motif under these conditions.

610 Bioconjugate Chem., Vol. 14, No. 3, 2003

Hoshika et al.

Figure 4. Quantitative EMSA to detect binding of TFO 13 (lanes 3-12) and TFO 14 (lanes 13-21) to DNA target. DNA concentration: 10 nM. TFO concentrations are indicated.

Figure 5. Specific inhibition of EcoT14I cleavage by the TFOs 8 (lanes 3-9), 13 (lanes 10-16), or 12 (lanes 17-23). Experimental conditions are described in Experimental Procedures.

To examine the effect of the composition and the length of a third strand on the alternate-strand triplex formation, the ability of the 3′-3′-linked TFO 13 consisting of T and G, the 3′-3′-linked TFO 14 comprised of A and G, the 3′-3′-linked TFO 15 composed of 16 bases, and the 3′-3′-linked TFO 16 consisting of 12 bases to bind to the DNA 17 was examined (Figure 4). The Kd of the TFO 13 to the DNA 17 was 41 ( 6 nM, whereas that of the TFO 14 was 17 ( 3 µM. It has been reported that the sequences of the junction regions of the alternate-strand triplex critically influence the stabilities of the triplexes. It was found that the TFO consisting of T and G stabilizes the alternate-strand triplex more efficiently than that composed of A and G. It has been reported that the presence of a series of alternating d(GA) repeats within the sequence facilitates formation of parallel-stranded homoduplexes with G:G and A:A base pairs (28-30). Thus, the self-association of the 3′-3′-linked TFO 14 may be partially responsible for the difference of the Kd values between the TFO 13 and 14. On the other hand, the Kd values of the TFOs 15 and 16 to the DNA 17 were 340 ( 10 nM and more than 100 µM, respectively. Thus, the TFO composed of more than 16 bases is required for the stable alternate-strand triplex formation by the purine motif. To investigate the base recognition ability of the 3′-3′linked TFOs, we tested the ability of the 3′-3′-linked TFO 13 to bind to the DNA target 18 involving one C:G base pair that cannot form reverse Hoogsteen hydrogen bonding with the G residue in the TFO 13. The Kd value of the TFO 13 to the DNA target 18 was more than 100 µM. Thus, it was revealed that the 3′-3′-linked TFOs have high base recognition ability. This result also indicates

that the 3′-3′-linked TFO 13 binds simultaneously to both the adjacent oligopurine domains of the DNA target 17 to form the stable triplex. Inhibition of Restriction Endonuclease Cleavage via Triplex Formation. We next tested the ability of the 3′-3′-linked TFOs to inhibit cleavage of a target DNA by a restriction endonuclease. The restriction enzyme EcoT14I recognizes the palindromic 6 bp sequence 5′CCTAGG-3′/3′-GGATCC-5′ and cleaves the two strands in a symmetric way (31). The DNA 17 contains the EcoT14I recognition site at its center and also involves oligopurine sequences, one on each strand, that can be recognized by the third strand. The unlinked decamer 8, the 3′-3′-linked TFO 13, and the 3′-3′-linked TFO 12 with the 2-anthraquinonecarbonyl group were used as the third strands. A series of experiments was carried out at 37 °C in a 50 mM Tris-HCl buffer (pH 7.5) containing 25 mM MgCl2 and 1 mM DTT. Increasing amounts of the third strand were added to the DNA 17, labeled at one 5′-end by 32P, and the solutions were preincubated at 37 °C for 30 min before addition of the restriction enzyme. After being incubated at 37 °C for 2 h, the products were analyzed by polyacrylamide gel electrophoresis under denaturing conditions (Figure 5). Densities of radioactivity of the gel were visualized by a Bio-imaging analyzer. As shown in Figure 6, the cleavage was inhibited in a dose-dependent manner when the 3′-3′-linked TFO 12 or 13 was used as the third strand. The 50% inhibitory concentrations (IC50s) of the 3′-3′-linked TFOs 12 and 13 were 0.90 and 0.48 µM, respectively. On the other hand, inhibition of the cleavage by the enzyme was not observed up to 200 µM with the TFO 8 as the third strand.

Alternate-Strand Triple Helix

Figure 6. Specific inhibition of EcoT14I cleavage by triplex formation: plots of percentage remaining DNA versus the concentration of TFO in the reaction mixture. The symbols on the plot indicate the TFOs 13 (O), 12 (4), and 8 (0).

To show that the inhibition was due to the interaction of the TFOs with the target DNA rather than with the restriction enzyme, a parallel experiment was carried out with the DNA 19, which contains one restriction site for EcoT14I but no sequence fully complementary to that of the TFO 13. It was found that the TFO 13 does not inhibit the cleavage of the DNA 19 by the enzyme at all (data not shown). Therefore, the inhibition of the cleavage was not due to the TFOs binding to the enzyme but to triplex formation at the binding site of the enzyme. CONCLUSION

In this paper, we describe the synthesis of the 3′-3′linked TFOs that can form the antiparallel triplexes with the DNA target by reverse Hoogsteen hydrogen bonds. Stability of the alternate-strand triplexes between these TFOs and the target duplex DNAs was investigated by the EMSA. It was found that the alternate-strand triplexes were significantly stabilized by linking the TFO fragments with the pentaerythritol linker. To our knowledge, this is the first example of the alternate-strand triplex composed of only the purine motif. In addition, we found that in contrast to the alternate-strand triplexes composed of the pyrimidine motif, the terminal ammonium ion of the aminobutyl-linker and the intercalator of the TFOs did not contribute to the stability of the alternate-strand triplex comprised of the purine motif. We tested the ability of the 3′-3′-linked TFOs to inhibit cleavage of the DNA target 17 by the restriction enzyme EcoT14I. It was revealed that the 3′-3′-linked TFO 12 and 13 inhibited the cleavage by the enzyme more effectively than the unlinked decamer 8. Thus, the TFOs linked with pentaerythritol may be useful as antigene oligonucleotides to the duplex DNAs, which have the alternating oligopyrimidine-oligopurine sequences. EXPERIMENTAL PROCEDURES

General Remarks. NMR spectra were recorded at 270 MHz (1H) and at 100 MHz (13C). Chemical shifts (δ) and coupling constants (J) are reported in parts per million downfield from TMS and in Hz, respectively. Mass spectra were obtained by fast atom bombardment (FAB) method. Thin-layer chromatography was done on Merck coated plates 60F254. The silica gel used for column chromatography was Merck silica gel 5715. 2,2-Bis-(4,4′-dimethoxytrityloxymethyl)-1,3-propanediol (2). After successive coevaporation with dry pyridine, pentaerythritol 1 (2.7 g, 20 mmol) was dissolved

Bioconjugate Chem., Vol. 14, No. 3, 2003 611

in dry pyridine (200 mL). DMTrCl (13.6 g, 40 mmol) was added in four portions, and stirred for 18 h at room temperature. EtOH (20 mL) was added to the solution, which was stirred at room temperature. After 30 min, the mixture was concentrated in vacuo and was taken in EtOAc, which was washed with aqueous NaHCO3 (saturated) and brine, dried (Na2SO4), and concentrated. The residue was purified by column chromatography (SiO2, EtOAc) to give 2 (3.9 g, 27% as a yellow foam): 1H NMR (CDCl3) δ 7.37-6.78 (m, 26H), 3.78 (s, 12H), 3.783.74 (m, 4H), 3.31 (s, 4H), 2.19-2.14 (t, 2H, J ) 6.6). 13C NMR (CDCl3) δ 45.4, 55.2, 63.2, 65.5, 86.2, 113.1, 126.7, 127.8, 128.0, 130.0, 135.5, 144.5, 158.3. HRMS calcd for C47H48O8Na (M + Na+), 763.3247; found, 763.3235. Anal. Calcd for C47H48O8: C, 76.19; H, 6.53. Found: C, 75.95; H, 6.65. 2-Benzoyloxymethyl-1,3-bis-O-(4,4′-dimethoxytrityl)-2-hydroxymethyl-1,3-propanediol (3). After successive coevaporation with dry benzene, 2 (400 mg, 0.54 mmol) was dissolved in dry benzene (20 mL). Bu2SnO (147 mg, 0.59 mmol) was added to the solution, which was heated under reflux for 6 h. The mixture was evaporated and dissolved in dry benzene (20 mL). BzCl (69 µL, 0.59 mmol) was added to the solution, which was stirred at room temperature. After 90 min, aqueous NaHCO3 (10 mL, saturated) was added to the solution, which was stirred at room temperature. After 15 h, the mixture was diluted with EtOAc. The organic layer was washed with aqueous NaHCO3 (saturated) and brine, dried (Na2SO4), and concentrated. The residue was purified by column chromatography (SiO2, 65% EtOAc in hexane) to give 3 (451 mg, 98% as a white foam): 1H NMR (CDCl3) δ 7.76-6.71 (m, 31H), 4.45 (s, 2H), 3.72 (s, 12H), 3.72-3.64 (m, 2H), 3.33 (s, 4H), 2.34 (bs, 1H). 13 C NMR (CDCl3) δ 45.1, 55.1, 62.0, 63.8, 64.6, 86.2, 113.0, 126.6, 127.7, 127.9, 128.2, 129.6, 129.7, 130.0, 132.9, 135.5, 144.6, 158.3, 166.2. HRMS calcd for C54H52O9Na (M + Na+), 867.3509; found, 867.3486. Anal. Calcd for C54H52O9: C, 76.76; H, 6.20. Found: C, 76.67; H, 6.25. 2-Benzoyloxymethyl-1,3-bis-O-(4,4′-dimethoxytrityl)-2-(O-succinyl)oxymethyl-1,3-propanediol (4). Compound 3 (300 mg, 0.36 mmol) was dissolved in dry CH3CN (5 mL). DMAP (44 mg, 0.36 mmol), Et3N (176 µL, 1.26 mmol), and succinic anhydride (126 mg, 1.26 mmol) was added to the solution, which was stirred at room temperature. After 3 days, aqueous NaHCO3 (5 mL, saturated) was added to the mixture, which was concentrated in vacuo. The residue was taken in EtOAc, which was washed with aqueous NaHCO3 (saturated) and brine, dried (Na2SO4), and concentrated. The residue was purified by column chromatography (SiO2, EtOAc) to give 4 (257 mg, 75% as a white foam): 1H NMR (CDCl3) δ 7.74-6.69 (m, 31H), 4.38 (s, 2H), 4.23 (s, 2H), 3.71 (s, 12H), 3.30 (s, 4H), 2.46-2.41 (m, 4H). 13C NMR (CDCl3) δ 28.6, 28.7, 44.0, 55.1, 60.2, 63.1, 63.3, 85.9, 112.9, 126.5, 127.6, 128.0, 128.2, 129.5, 129.7, 130.0, 132.9, 135.5, 144.6, 158.2, 165.9, 171.4, 176.5. HRMS calcd for C58H56O12Na (M + Na+), 967.3669; found, 967.3650. Anal. Calcd for C58H56O12: C, 73.71; H, 5.97. Found: C, 73.49; H, 6.18. Synthesis of CPG 5. Aminopropyl controlled pore glass (448 mg, 40 µmol, 90.4 µmol/g, CPG Inc.) was added to a solution of 4 (150 mg, 0.159 mmol) and 1-ethyl-3(3-dimethylaminopropyl)carbodiimide hydrochloride (31 mg, 0.159 mmol) in dry DMF (4 mL), and the mixture was kept at room temperature for 3 days. After the resin was washed with pyridine, 5 mL of a capping solution (0.5 M DMAP in pyridine:Ac2O ) 9:1) was added, and the whole was kept at room temperature for 15 h. The

612 Bioconjugate Chem., Vol. 14, No. 3, 2003

resin was washed with EtOH and acetone and was dried under vacuum. The loaded 4 to the solid support is 25 µmol/g from the calculation of released dimethoxytrityl cation by a solution of 70% HClO4:EtOH (3:2, v/v). Synthesis of TFOs. TFOs were synthesized on a DNA synthesizer (Applied Biosystem Model 392) by the phosphoramidite method. Each TFO linked to the resin was treated with concentrated NH4OH at 55 °C for 16 h, and the released TFO was purified by denaturing 20% polyacrylamide gel electrophoresis (20% PAGE) with 7 M urea at 600 V for 8 h. The TFO was visualized by UV shadowing and eluted from crushed gel slices by incubation at room temperature in 0.1 M Tris buffer (pH 7.0) and 1 mM EDTA for 15 h. The TFO was purified on a Sep-Pak C18 cartridge to give the deprotected TFO 10 (13.6), 12 (10.8), 13 (6.6), 14 (9.9), 15 (6.7), and 16 (7.1). The yields are indicated in parentheses as OD units at 260 nm starting from 0.25 µmol scale. Acetylation of TFO 10. A solution containing TFO 10 (5.0 OD units at 260 nm) and Ac2O (10 µL) in 0.2 M HEPES buffer (pH 7.2, 1.0 mL) was kept for 5 h at room temperature. Concentrated NH4OH (2.0 mL) was added to the mixture, and the whole was kept overnight at 4 °C. The solvent was removed in vacuo, and the residue was purified by HPLC with a C18 column to give TFO 11 (3.7 OD units at 260 nm). Matrix-Assisted Laser Desorption/Ionization Timeof-Flight Mass Spectrometry. Spectra were obtained on a Voyager Elite reflection time-of-flight mass spectrometry (PerSeptive Biosystems, Inc., Framingham, MA) equipped with a nitrogen laser (337 nm, 3 ns pulse). 3-Hydroxypicolinic acid (HPA), dissolved in H2O to give a saturated solution at room temperature, was used as the matrix. TFO 10: calculated mass, 6633.31; observed mass, 6634.87. TFO 11: calculated mass, 6675.35; observed mass, 6675.85. TFO 12: calculated mass, 6867.57; observed mass, 6868.89. TFO 13: calculated mass, 6519.17; observed mass, 6520.42. TFO 14: calculated mass, 6591.27; observed mass, 6592.76. TFO 15: calculated mass, 5252.37; observed mass, 5252.74. TFO 16: calculated mass, 3935.54; observed mass, 3936.74. Electrophoretic Mobility Shift Assays. DNA (10 nM) labeled at one 5′-end by 32P was incubated in the presence of various amounts of TFO in a 20 mM TrisHCl buffer (pH 7.5) containing 10 mM MgCl2 (total 20 µL) at 70 °C for 5 min and then gradually cooled to 4 °C. The reaction mixture was supplemented with 4 µL of a 50% glycerol solution containing bromophenol blue. The reaction was analyzed by electrophoresis on 20% nondenaturing polyacrylamide gel (39:1 acrylamide:bisacrylamide) using a TB buffer (45 mM Tris borate, pH 8, 10 mM MgCl2) at 60-80 mV for 24-36 h at 4 °C. After drying, density of radioactivity of the gel was visualized by a Bio-imaging Analyzer (Bas 2000, Fuji, Co. Ltd.). Enzymatic Assays. Enzymatic assays were performed in a 50 mM Tris-HCl buffer (pH7.5) containing 25 mM MgCl2, and 1 mM DTT. Various amounts of the third strands were added to the duplex (1 pmol), labeled at one 5′-end by 32P, and the solutions were preincubated at 37 °C for 30 min. Twenty units of EcoT14I (Takara Shuzo Co., Ltd.) was added to each solution, and the mixture (total 10 µL) was incubated at 37 °C for 2 h. One aliquot (1.5 µL) of the reaction mixture was separated and added to a loading solution (8 M urea, 0.1% xylene cyanol, 0.1% bromophenol blue, 5 µL). The solution was analyzed by electrophoresis on 20% polyacrylamide gel containing 7 M urea. Densities of radioactivity of the gel were visualized by a Bio-imaging analyzer (Bas 2000, Fuji Co., Ltd).

Hoshika et al. ACKNOWLEDGMENT

This work was supported in part by a Grant-in-Aid for Encouragement of Young Scientists from the Ministry of Education, Science, Sports, and Culture of Japan and a grant from the “Research for the Future” Program of the Japan Society for the Promotion of Science (JSPSRFTF97I00301). We thank Y. Misawa (Hokkaido University) for technical assistance. LITERATURE CITED (1) Thuong, N. T., and He´le`ne, C. (1993) Sequence specific recognition and modification of double-helical DNA by oligonucleotides. Angew. Chem., Int. Ed. Engl. 32, 666-690. (2) Moser, H. E., and Dervan, P. B. (1987) Sequence specific cleavage of double helical DNA by triple helix formation. Science 238, 645-650. (3) Rajagopal, P., and Feigon, J. (1989) Triple-strand formation in the homopurine:homopyrimidine DNA oligonucleotides d(G-A)4 and d(T-C)4. Nature 339, 637-640. (4) Beal, P. A., and Dervan, P. B. (1991) Second structural motif for recognition of DNA by oligonucleotide-directed triple-helix formation. Science 251, 1360-1363. (5) Pilch, D. S., Levenson, C., and Shafer, R. H. (1991) Structure, stability, and thermodynamics of a short intermolecular purine-purine-pyrimidine triple helix. Biochemistry 30, 60816088. (6) Singleton, S. F., and Dervan, P. B. (1992) Influence of pH on the equilibrium association constants for oligodeoxyribonucleotide-directed triple helix formation at single DNA sites. Biochemistry 31, 10995-11003. (7) Beal, P. A., and Dervan, P. B. (1992) Recognition of double helical DNA by alternate strand triple helix formation. J. Am. Chem. Soc. 114, 4976-4982. (8) Jayasena, S. D., and Johnston, B. H. (1992) Oligonucleotidedirected triple helix formation at adjacent oligopurine and oligopyrimidine DNA tracts by alternate strand recognition. Nucleic Acids Res. 20, 5279-5288. (9) Jayasena, S. D., and Johnston, B. H. (1992) Intramolecular triple-helix formation at (PunPyn)‚(PunPyn) tracts: recognition of alternate strands via Pu‚PuPy and Py‚PuPy base triplets. Biochemistry 31, 320-327. (10) Balatskaya, S. V., Belotserkovskii, B. P., and Johnston, B. H. (1996) Alternate-strand triplex formation: modulation of binding to matched and mismatched duplexes by sequence choice in the Pu-Pu-Py Block. Biochemistry 35, 1332813337. (11) Washbrook, E., and Fox, K. R. (1994) Alternate-strand DNA triple-helix formation using short acridine-linked oligonucleotides. Biochem. J. 301, 569-575. (12) Washbrook, E., and Fox, K. R. (1994) Comparison of antiparallel A.AT and T. AT triplets within an alternate strand DNA triple helix. Nucleic Acids Res. 22, 3977-3982. (13) De Bizemont, T., Sun, J.-S., Garestier, T., and He´le`ne, C. (1998) New junction models for alternate-strand triple-helix formation. Chem. Biol. 5, 755-762. (14) Brodin, P., Sun, J.-S., Mouscadet, J.-F., and Auclair, C. (1999) Optimization of alternate-strand triple helix formation at the 5′-TpA-3′ and 5′-ApT-3′ junctions. Nucleic Acids Res. 27, 3029-3034. (15) Horne, D. A., and Dervan, P. B. (1990) Recognition of mixed-sequence duplex DNA by alternate-strand triple-helix formation. J. Am. Chem. Soc. 112, 2435-2437. (16) Ono, A., Chen, C. N., and Kan, L. S. (1991) DNA triplex formation of oligonucleotide analogues consisting of linker groups and octamer segments that have opposite sugarphosphate backbone polarities. Biochemistry 30, 9914-9921. (17) McCurdy, S., Moulds, C., and Froehler, B. (1991) Deoxyoligonucleotides with inverted polarity: synthesis and use in triple-helix formation. Nucleosides Nucleotides 10, 287-290. (18) Froehler, B. C., Terhorst, T., Shaw, J. P., and McCurdy, S. N. (1992) Triple-helix formation and cooperative binding by oligodeoxynucleotides with a 3′-3′ internucleotide junction. Biochemistry 31, 1603-1609.

Alternate-Strand Triple Helix (19) Asseline, U., and Thuong, N. T. (1993) Oligonucleotides tethered via nucleic bases. A potential new set of compounds for alternate strand triple-helix formation. Tetrahedron Lett. 34, 4173-4176. (20) Asseline, U., and Thuong, N. T. (1994) 5′-5′ Tethered oligonucleotides via nucleic bases: a potential new set of compounds for alternate strand triple-helix formation. Tetrahedron Lett. 35, 5221-5224. (21) Asseline, U., Roig, V., and Thuong, N. T. (1998) Modified oligonucleotides with 5′-5′ interbase semirigid junction for alternate strand triplex formation. Tetrahedron Lett. 39, 8991-8994. (22) Zhou, B.-W., Marchand, C., Asseline, U., Thuong, N. T., Sun, J.-S., Garestier, T., and He´le`ne, C. (1995) Recognition of alternating oligopurine/oligopyrimidine tracts of DNA by oligonucleotides with base-to-base linkages. Bioconjugate Chem. 6, 516-523. (23) Ueno, Y., Ogawa, A., Nakagawa, A., and Matsuda, A. (1996) Nucleosides and nucleotides. 162. Facile synthesis of 5′-5′-linked oligodeoxyribonucleotides with the potential for triple-helix formation. Bioorg. Med. Chem. Lett. 23, 28172822. (24) De Napoli, L., Messere, A., Montesarchio, D., Pepe, A., Piccialli, G., and Varra, M. (1997) Synthesis and triple helix formation by alternate strand recognition of oligonucleotides containing 3′-3′ phosphodiester bonds. J. Org. Chem. 62, 9024-9030.

Bioconjugate Chem., Vol. 14, No. 3, 2003 613 (25) Ueno, Y., Mikawa, M., Hoshika, S., and Matsuda, A. (2001) Alternate-strand triple-helix formation by the 3′-3′-linked oligodeoxynucleotides with the anthraquinonyl group at the junction point. Bioconjugate Chem. 12, 635-642. (26) Wang, G., Seidman, M. M., and Glazer, P. M. (1996) Mutagenesis in mammalian cells induced by triple helix formation and transcription-coupled repair. Science 271, 802-805. (27) Cogoi, S., Rapozzi, V., Quadrifoglio, F., and Xodo, L. (2001) Anti-gene effect in live cells of AG motif triplex-forming oligonucleotides containing an increasing number of phosphorothioate linkages. Biochemistry 40, 1135-1143. (28) Rippe, K., Fritsch, V., Westhof, E., and Jovin, T. M. (1992) Alternating d(G-A) sequences form a parallel-stranded DNA homoduplex. EMBO J. 11, 3777-3786. (29) Huertas, D., Bellsolell, L., Casasnovas, J. M., Coll, M., and Azorin, F. (1993) Alternating d(GA)n DNA sequences form antiparallel stranded homoduplexes stabilized by the formation of G‚A base pairs. EMBO J. 12, 4029-4038. (30) Noonberg, S. B., Franc¸ ois, J.-C., Garestier, T., and He´le`ne, C. (1995) Effect of competing self-structure on triplex formation with purine-rich oligodeoxynucleotides containing GA repeats. Nucleic Acids Res. 23, 1956-1963. (31) Mise, K., and Nakajima, K. (1985) Purification of a new restriction endonuclease, StyI, from Escherichia coli carrying the hsd+ miniplasmid. Gene 33, 357-61.

BC0256493