DNA Photocleavage by DNA and DNA−LNA Amino Acid−Dye

A plasmid of this gene (P-CMV-mdm2) was used as the DNA target; this ...... M. M. (2008) Targeted gene knock in and sequence modulation mediated by a ...
0 downloads 0 Views 1MB Size
616

Bioconjugate Chem. 2010, 21, 616–621

DNA Photocleavage by DNA and DNA-LNA Amino Acid-Dye Conjugates Adva Biton, Aviva Ezra, Jana Kasparkova, Viktor Brabec, and Eylon Yavin* Department of Medicinal Chemistry, The Institute for Drug Research, The School of Pharmacy, The Hebrew University of Jerusalem, Hadassah Ein-Karem, Jerusalem 91120, Israel, and Institute of Biophysics, Academy of Sciences of the Czech Republic, v.v.i., 61265 Brno, Czech Republic. Received August 26, 2009; Revised Manuscript Received December 28, 2009

DNA photocleavage by triplex forming oligonucleotides (TFO) has potential implications in both biotechnology and medicine. We have synthesized a series of homopurine DNA and DNA/LNA 14-mers to which an amino acid (glycine or L-tryptophan) and a cyanine dye are covalently linked. Two cyanine dyes were examined that include a quinolinium ring linked to a benzothiazolium ring through a monomethine (TO1) or trimethine (TO2) linker. The 14-mer sequence was chosen to target mdm2, a ubiquitin ligase (E3) that regulates p53 by promoting its ubiquitylation and proteosomal degradation. Such inhibition has been previously proposed as a therapeutic approach to target wild-type p53-expressing cancers. To examine whether our TFO conjugates photocleave the mdm2 target, we incubated the various conjugates with the mdm2 plasmid and irradiated the samples with visible light. We show that only the TFO with the complementary sequence and with an intervening L-tryptophan leads to the linearization of the plasmid after a short irradiation time (10 min) exciting the dye (λmax(TO1) ) 500 nm and λmax(TO2) ) 630 nm) with visible light. Furthermore, the photoreactivity is more pronounced for the LNA/ DNA conjugate, an observation that is consistent with improved hybridization to the DNA target. Sequence specificity of the photoreaction is further corroborated on a synthetic 44-mer duplex containing the TFO site. Evidence for a ROS-dependent mechanism is also given and discussed.

INTRODUCTION The selective photocleavage of DNA holds great therapeutic potential as means of down-regulating gene expression in a controlled and persistent manner (1-4). A common and attractive approach depends on light activation of a DNA-dye conjugate, leading to the formation of reactive oxygen species (ROS) that are localized at the target site (2, 3, 5). Such an approach merits from the control in localizing the damage at a given area (e.g., tissue) and at a timely fashion. Several studies have shown that such site-specific selectivity is achievable. Notable are extensive studies with psoralenconjugates, that upon UV irradiation lead to stable cross-links on the target duplex DNA (6-24). Some ODN-conjugates have shown site-specific photoreactivity upon irradiation in the visible region (25-32), and only very few reports have applied dyes that have relevant wavelengths for in vivo applications (i.e. over 600 nm) (33-35). For example, a lutetium (III) texaphyrin complex tethered to a 2′-OMe RNA analogue was shown to selectively photocleave a complementary DNA strand after sample irradiation (λ ) 732 nm) followed by piperidine treatment, resulting in DNA strand breaks primarily at guanine residues (34). Such reactivity is a consequence of singlet oxygen generation that preferably oxidizes guanine, the base with the lowest oxidation potential of all four natural DNA bases. Piperidine is then required in order to promote DNA scission at oxidized sites. The use of visible light to achieve direct DNA strand breaks would be advantageous, since (1) there would be no toxicity due to irradiation (as opposed to UV) and (2) DNA damage would be more deleterious (strand break as opposed to nucleobase oxidation). Thiazole orange (TO) is a highly fluorescent intercalating dye and an efficient singlet oxygen sensitizer (36). TO has negligible fluorescence in solution, and obtains intense fluorescence when bound to nucleic acids. Kelley and co* [email protected], tel: 972-2-6758692, fax: 972-2-6757076.

workers have shown that, upon tethering TO to a dipeptide having either L-tryptophan (W) or L-tyrosine (Y), photoactivation of the dye (at 500 nm) resulted in the generation of a stable peroxide on the aromatic ring promoting nonselective DNA cleavage (37, 38). It was also shown that Y and W were required for the observed DNA cleavage (as determined by supercoiled plasmid nicking) and that replacing these amino acids with glycine or L-phenylalanine resulted in a peptide-dye conjugate that had no apparent photoinduced activity toward plasmid DNA. Inspired by the photoreactivity of TO-peptides discovered by the Kelley group, we decided to explore the possibility of tethering TO-tryptophan to a TFO in order to achieve frank DNA strand breaks after visible light irradiation of the conjugate in the presence of the target plasmid DNA. In addition to monomethine TO-conjugates, we synthesized a series of DNA conjugates that include a cyanine dye with a trimethine linker (TO2) (39), as this dye is photoexcited at a wavelength (630 nm) that is relevant for tissue penetration (i.e., photodynamic therapy) in an in vivo setting. Furthermore, we explored the outcome of introducing LNA bases in the TFO in order to improve the hybridization of such TFOs to the targeted DNA (40-46). We chose a TFO consisting of the homopurine sequence 5′-AAAGGAAAGGGAAA to target the Mdm2 oncogene. Mdm2 acts as an inhibitor of p53 by accelerating its intracellular degradation. As Mdm2 is overexpressed in many cancer cells, it is an attractive anticancer target (47-49). Herein, we report on the synthesis of the TFO conjugates and their photoreactivity toward the mdm2 gene in its DNA plasmid form and in a synthetic DNA duplex containing the TFO binding site.

EXPERIMENTAL PROCEDURES Irradiation with Plasmid PCMV-MDM2. Photoinduced DNA cleavage experiments were done using the MDM2 plasmid

10.1021/bc900372h  2010 American Chemical Society Published on Web 03/26/2010

DNA Photocleavage by DNA and DNA-LNA Conjugates

(8427 bp’s) containing the target sequence that was incubated with the various DNA conjugates overnight at 37 °C. For DNA conjugates, the following buffer was used for incubation: 2 mM EDTA, 20 mM Tris, and 200 mM NaCl (pH ) 7.4). For DNA/ LNA conjugates, the following buffer was used for incubation: 10 mM NaH2PO4 and 50 mM NaCl (pH ) 7.4). Each of the conjugates were irradiated (in a dark room) for the indicated time with visible light. The light source used was a 150 W halogen lamp. This lamp (mrc) has a fiber optic that allows us to aim the light source onto the sample. Samples were placed in an eppendorf tube and the light source (from the filament) was about 1 cm away from the sample. Visible light was filtered with a cutoff filter (passing light between 400 to 600 nm for TO1 and 500 to 750 nm for TO2). Irradiation with Synthetic 44-mer Duplex. The 44-mer duplex DNA as well as ssDNA markers were purchased from Synthezza (Jerusalem, Israel) and used as received. These sequences are as follows: 44-mers (with TFO target underlined), 3′ TCGGTCCGAAAGTAGTTTCCTTTCCCTTTATGATAGTCTAAACA-5′; and its complementary sequence, 5′-AGCCAGGCTTTCATCAAAGGAAAGGGAAATACTATCAGATTTGT-3′. 23-mers: 3′-TTCCCTTTATGATAGTCTAAACA-5′ and 5′-AAGGGAAATACTATCAGATTTGT-3′. 20-mers: 3′-CCTTTATGATAGTCTAAACA-5′ and 5′-GGAAATACTATCAGATTTGT-3′. 17-mer: 3′-CCTTTATGATAGTCTAAACA-5′. 14-mer: 3′-TTATGATAGTCTAAACA-5′. The 44-mer duplex containing the 14 base targeting sequence at its center was incubated (37 °C, 16 h) with an equimolar concentration (100 µM) of conjugate 3a. The following buffer was used for incubation: 10 mM Tris, 100 mM NaCl, and 1 mM EDTA. All the samples were irradiated (in a dark room) for the indicated time (30 min) with visible light as described for plasmid irradiation. Gel Electrophoresis. Agarose (1%) gel electrophoresis was carried out using Tris-acetate-EDTA (TAE) buffer. Running conditions: 60 min at 100 V. Denaturing (15% Urea) PAGE (19:1 acrylamide/bis-acrylamide) was carried out in Tris-borateEDTA (TBE) buffer. Running conditions: 120 min at 100 V. The gels were stained with EB for 15 min and washed with DDW. Tm Measurements. All thermal denaturation studies were carried out using quartz cuvettes with an optical path length of 1 cm on a Cary 300 UV/vis spectrophotometer interfaced with a computer for data collection and analysis. The temperature was increased and decreased from 20 to 80 °C at the rate of 1 °C/min. Samples were heated at 90 °C for 5 min and cooled down to room temperature for several hours prior measuring the melting profile of the dsDNA (44-mer). For triplex formation, the LNA-TFO (conjugate 3a) was incubated with the dsDNA at 37 °C for 16 h. Thermal denaturation experiments were carried out on a sample containing 1 µM duplex and same concentration of TFO for the triplex study. All samples were prepared in 10 mM Tris, 100 mM NaCl, and 1 mM EDTA. The melting temperature was determined by plotting the first derivative of the absorbance versus temperature profile and is accurate to (1 °C. Solid-Phase Synthesis of DNA and DNA/LNA Conjugates. The DNA (or DNA/LNA) oligomers were synthesized on an automated DNA synthesizer (ABI 3400) using standard phosphoramidite chemistry removing the DMT group at the 5′ end. All subsequent coupling reactions were performed on the solid phase (CPG resin (Glen Research), 2 µmol scale). The 5′-OH end of the oligos were extended with an ethylenediamine spacer by reacting the 5′-OH group with carbonyldiimidazole (CDI) (50 µmol CDI in 1 mL CH3CN) for 1 h. After washing resin with CH3CN (3 × 1 mL), ethylenediamine (10

Bioconjugate Chem., Vol. 21, No. 4, 2010 617 Table 1 conjugate

% nickeda (form II)

% lineara (form III)

MALDI-TOF MS obs./calc. (M + 1)

1a 1b 1c 2a 2b 2c 3a 3b 3c

41.7 36 12 51.3 21.6 14.5 55.4 33 9.8

11.9

5063.4/5064.0 4932.0/4934.0 5063.7/5064.0 5116.4/5114.8 4831.8/4833.0 5116.2/5114.8 5234.5/5232.2 5103.0/5102.2 5234.3/5232.2

a

6.4 20.4

Values are for 30 min irradiation.

µmol) and diisopropylethylamine (DIEA, 12 µmol) were added, and the reaction vessel was shacked overnight at RT. After washing the resin with dioxane (3 × 1 mL), CH3CN (2 × 1 mL), and DCM (2 × 1 mL), coupling reactions with NPS (Onitrophenylsulphenyl)-protected amino acids (L-tryptophan and glycine) were carried out using 4 equiv NPS-a.a. (8 µmol), 4.8 equiv HOBT (9.6 µmol), 4.8 equiv (9.6 µmol) HATU, and 8 equiv (16 µmol) DIEA in dry N,N-dimethylformamide (DMF, 1 mL) for 3 h. The NPS protecting group was removed with 2% dichloroacetic acid (DCA) in thioacetamide (TAA) (dissolved in 1 mL N-methylpyrrolidone ((NMP)) for 15 min. The free amine was finally coupled to the carboxy functionalized thiazole orange (TO1 or TO2, 4 equiv (8 µmol) under conditions as described for amino acid coupling. DNA and DNA/LNA conjugates were removed from the resin by treating solid support with a 33% solution of NH4OH for 1-2 h at room temperature. The supernatant was removed and incubated at 37 °C for 24 h to ensure the removal of all protecting groups from DNA and LNA bases. The NH4OH solution was dried under vacuum and the resulting solid was purified by RP-HPLC (Shimatzu LC2010) using a semipreparative C18 reverse-phase column (Phenomenex, Jupiter 300 A) at a flow rate of 3.5 mL/min. Conditions for HPLC separation of DNA and DNA/LNA conjugates are described in detail in Supporting Information. Mobile phase A was either 50 mM NH4OAc (pH 5.5) or 50 mM TEAA (pH ) 7) with 5% CH3CN; mobile phase B was either CH3CN or MeOH. HPLC purified DNA and DNA/LNA conjugates were characterized by MALDITOF MS (see Table 1).

RESULTS Synthesis of DNA and DNA/LNA-Amino Acid-Dye Conjugates. Scheme 1 shows the synthetic strategy for the synthesis of DNA and DNA/LNA-amino acid-dye conjugates with two cyanine dyes having either a monomethine linker (TO1) or a trimethine linker (TO2). Upon photoexcitation of such dye conjugates, the following processes are expected: (1) generation of singlet oxygen and/or peroxide upon photoactivation according to the choice of the intervening amino acid and (2) selective cleavage of the target DNA by conjugating a complementary TFO strand to the amino acid-dye moiety. The possible photochemically induced reactions are depicted in Scheme 2. Synthesis of cyanine dyes (TO1 and TO2) with pendant carboxylic acid moieties was accomplished according to the literature (39). Initially, the oligomer (14-mer homopurine, vide post) was synthesized on an automated DNA/RNA synthesizer on a standard CPG solid support. All subsequent reactions were carried out on the solid support including linker extension, as well as amino acid and dye conjugations. NPS protected amino acids (glycine and L-tryptophan) were used for all conjugates. As the removal of NPS is achievable at mild reducing acidolysis conditions (50), this type of chemistry should be suitable for peptide-ODN synthesis. Indeed, we have

618 Bioconjugate Chem., Vol. 21, No. 4, 2010

Biton et al.

Scheme 1. Solid-Phase Synthesis of DNA and DNA/LNA-Amino Acid-Dye Conjugates

Figure 1. Photocleavage activity of conjugates 1a-1c on the PCMVMDM2 (8427 bp) plasmid in a buffer consisting of 2 mM EDTA, 20 mM Tris, and 200 mM NaCl (pH ) 7.4) as determined by agarose gel electrophoresis. Solutions contained 2 µg plasmid and 10 µM TFO. Lanes 1-4: plasmid + conjugate 1a (with L-tryptophan) after 15, 20, and 30 min of irradiation and in the dark, respectively. Lanes 5-6: plasmid + conjugate 1b (scrambled) after 30 min of irradiation and in the dark. Lanes 7-8: plasmid + conjugate 1c (with glycine) in the dark and after 30 min of irradiation. Lane 9: plasmid after 30 min irradiation.

Figure 2. Oxygen dependence of photocleavage reaction of conjugates 1a and 1b with PCMV-MDM2 plasmid as determined by agarose gel electrophoresis. Solutions contained 2 µg plasmid and 10 µM TFO. Lanes 1 and 2, plasmid in the dark and after 30 min of irradiation; lanes 3-5, plasmid + conjugate 1a without inhibitor, with NaN3 (1 mM), and with Trolox (0.1 mM), respectively; lanes 6-8, plasmid + conjugate 1b with NaN3 (1 mM), without inhibitor and with Trolox (0.1 mM), respectively. The same buffer was used as in Figure 1. Scheme 2. Possible Photochemical Pathways of DNA and DNA/LNA-Amino Acid-Dye Conjugates

found that using NPS-protected amino acids is superior to that of Fmoc protected amino acids (based on HPLC analysis; data not shown). Finally, DNA and DNA/LNA conjugates were removed from the solid support by treating resins with a 33% solution of NH4OH for 1-2 h. Subsequently, base protecting groups were removed by incubating the ammonia solution at 37 °C for 24 h. All DNA conjugates were then HPLC purified on a reverse phase C18 column and analyzed by Maldi-TOF MS (see Supporting Information for HPLC chromatograms). Choice of Amino Acid and DNA Sequence. The introduction of L-tryptophan in close proximity to the dye is expected to lead to its oxidation via singlet oxygen sensitization yielding a peroxide on the indole ring (37, 38). Such an ROS is anticipated to cause direct DNA strand breaks. Of all the aromatic amino acids that exert such reactivity (Trp, Tyr, and His), L-tryptophan has been shown to yield the highest peroxide levels (51). A 14-mer homopurine sequence (5′-AAAGGAAAGGGAAA-3′) was chosen as a triplex-forming oligonucleotide (TFO) that has the propensity of forming a triplex with the sense strand of the mdm2 gene in an antiparallel fashion (4). A plasmid of this gene (P-CMV-mdm2) was used as the DNA target; this plasmid includes the 14-mer TFO target. Thus, the “wild-type” TFO includes an intervening Ltryptophan between the ODN and the dye (conjugate 1a, Scheme 1). As a comparison, we chose to introduce glycine at this position, as it is not expected to react with singlet oxygen, and in turn, such a conjugate should exclusively yield singlet oxygen upon photoactivation (conjugate 1b, Scheme 1). In addition, we synthesized a scrambled TFO sequence with an intervening L-tryptophan as a control conjugate that potentially may form a

peroxide on the indole ring but is not expected to form a triplex with the target DNA (conjugate 1c, Scheme 1). Conjugates 1a-1c all contain TO1 that is photoexcited at 500 nm. As a comparison, the analogous conjugates (conjugates 2a-2c) were synthesized that include TO2; a dye that is excited at a longer wavelength, namely, 630 nm. Finally a set of conjugates were synthesized with TO1 that included several LNA’s as means to improve triplex formation (conjugates 3a-3c). Photocleavage of Plasmid DNA by DNA Conjugates. Prior to irradiation, DNA conjugates were incubated for 16 h with the plasmid DNA at 37 °C. In the case of the all-DNA TFO conjugates (1a-c, 2a-c) a 100-fold molar excess of TFO was used to ensure triplex formation (24). As shown on the agarose gel (Figure 1), visible light irradiation of DNA conjugate 1a with TO1 leads to some DNA relaxation of the supercoiled form (I) to the relaxed form (II) with a faint band corresponding to the linearized form (III) (lanes 1-3). Photoirradiation of the DNA conjugate with glycine (1b) results in some DNA relaxation of the plasmid with no indication of linearization (lane 8). Finally, the conjugate with the scrambled sequence (conjugate 1c) has no apparent effect on the target plasmid DNA (lanes 5-6). Clearly, the photoreaction is sequence-specific, as the conjugate with the scrambled sequence (conjugate 1c) does not lead to any detectable DNA photocleavage. The DNA conjugates (2a-2c) with the TO2 dye were next examined for their photoactivity toward the mdm2 plasmid. As shown in Figure 3, a similar behavior was seen with these conjugates (2a-2c) as for the previous set (conjugates 1a-1c). Thus, conjugate 2a (with L-tryptophan) leads to extensive nicking of the supercoiled plasmid (lanes 6 and 7) with some indication for linearized plasmid (form III). Conjugate 2c (scrambled) shows no photoactivity (lanes 2-4), similar to the behavior observed for conjugate 1c. Finally, conjugate 2b (with glycine) leads to some DNA relaxation (form II) (lanes 9-10) with no sign of the linear form (III). Photocleavage of Plasmid DNA by DNA/LNA Conjugates. Prior to irradiation, DNA/LNA conjugates (3a-3c) were incubated for 16 h with the plasmid DNA at 37 °C. With

DNA Photocleavage by DNA and DNA-LNA Conjugates

Figure 3. Photocleavage of PCMV-MDM2 plasmid by DNA conjugates with an appended cyanine dye TO2 as determined by agarose gel electrophoresis. Solutions contained 2 µg plasmid and 10 µM TFO. Lane 1, plasmid irradiated for 1 h (DNA only); lanes 2-4, plasmid + conjugate 2c (scrambled) in the dark and after 1 h and 30 min irradiation, respectively; lanes 5-7, plasmid + conjugate 2a (with L-tryptophan) in the dark and after 1 h and 30 min of irradiation, respectively; lanes 8-10, plasmid + conjugate 2b (with glycine) in the dark and after 1 h and 30 min of irradiation, respectively. The same buffer was used as in Figure 1.

Figure 4. Photocleavage of PCMV-MDM2 plasmid by TFO containing LNA conjugates 3a-3c (with TO1) in phosphate buffer (10 mM NaH2PO4, 50 mM NaCl, pH ) 7.4), as determined by agarose gel electrophoresis. Lane 1, plasmid irradiated for 30 min (DNA only); lanes 2-6, plasmid + conjugate 3a (with L-tryptophan) after 10, 15, 20, and 30 min irradiation and in the dark, respectively; lanes 7-9, plasmid + conjugate 3c after 15 and 30 min irradiation and in dark, respectively; lanes 10-3, plasmid + TFO conjugate 3b in the dark and after 10, 15, and 30 min irradiation, respectively.

the LNA/DNA hybrid (conjugates 3a-3c), only a 2-fold access was used as such TFOs have been shown to readily from a triplex at such molar ratios. In order to verify triplex formation by the LNA conjugates, a synthetic 44-mer duplex containing the 14 base targeting sequence at its center was incubated (37 °C, 16 h) with an equimolar concentration (100 µM) of conjugate 3a. Thermal denaturation profiles of the duplex and the triplex were carried out (see Supporting Information for thermal denaturation curves). A Tm value of 52 °C was determined by subtracting the triplex profile from that of the duplex profile. This value is similar (Tm ) 50 °C) to that reported for a homopyrimidine 15-mer TFO containing 6 LNA’s (52). The photoactivity of these set of conjugates is shown in Figure 4. Conjugate 3a is shown to cause substantial linearization of the mdm2 plasmid after a short irradiation period of 10 min (lane 2). No significant change in the amount of nicked and linearized plasmid is observed at longer periods of irradiation (lanes 3-5). The scrambled LNA/DNA conjugate (3c) has no apparent photoactivity, whereas the conjugate with glycine (3b) leads only to plasmid DNA relaxation (form II). Photocleavage of a Synthetic 44-mer Duplex DNA by DNA/LNA Conjugate 3a. To gain insight into the selectivity of DNA photocleavage by the LNA TFO (conjugate 3a), a synthetic 44-mer duplex containing the 14-mer target site of conjugate 3a was utilized as the DNA target. The 44-mer strands were chosen from the mdm2 plasmid with 15-mers flanking both sides of the central 14-mer target (underlined) as follows: 3′-TCGGTCCGAAAGTAGTTTCCTTTCCCTTTATGATAGTCTAAACA-5′,5′-AGCCAGGCTTTCATCAAAGGAAAGGGAAATACTATCAGATTTGT-3′. A denaturing polyacrylamide gel (subsequently stained with EtBr) was used to follow the photoreaction (Figure 5). In lanes 1-6, various lengths of ssDNA were added as DNA markers in order to estimate the length of the DNA photoproduct formation. As shown in lane 9, a new migrating band is seen above the faster migrating TFO (3a) after sample irradiation (30 min); a product that is not observed in the dark control (lane 10: TFO + 44-mer duplex DNA without irradiation). This band migrates similarly to the 17-mer (lane 5). Thus, this gel is indicative of a site-specific photoreaction leading to a distinct

Bioconjugate Chem., Vol. 21, No. 4, 2010 619

Figure 5. Photocleavage of a synthetic duplex DNA (a 44-mer containing the TFO targeting site at its center) by the TFO containing LNA conjugate 3a (with TO1) in phosphate buffer (10 mM NaH2PO4, 50 mM NaCl, pH ) 7.4), as determined by 15% denaturing polyacrylamide gel electrophoresis. Lanes 1-6: ssDNA markers with the same sequence composition as the 44-mer. 20- and 23-mers are from the 5′-end of the TC-rich 44-mer and from the 3′-end of the AG-rich 44mer, respectively. 14- and 17-mers are from the 5′-end of the TC-rich 44-mer. Lanes 7 and 8: 44-mer duplex DNA in the dark and after irradiation (30′), respectively. Lanes 9 and 10: 44-mer duplex DNA and the LNA conjugate 3a after 30′ irradiation and in the dark, respectively. Blue arrow indicates the formation of a new photoinduced DNA product. The gel was run in Tris-borate-EDTA (TBE) buffer for 120 min at 100 V and stained with ethidium bromide.

band that is in the range of 16-18 bases. It is not clear as to the fate of the remaining duplex (ca. 29 base pairs). We assume that this duplex is not well resolved and remains with the original 44-mer duplex at the upper part of the gel. Photocleavage Activity Is Oxygen Dependent. In order to verify the involvement of reactive oxygen species (ROS) in the DNA photocleavage reaction, two ROS scavengers were used. Inhibition of cleavage in the presence of singlet oxygen quencher NaN3 indicates the involvement of singlet oxygen as the reactive species (53) whereas a radical scavenger such as Trolox (37, 54) indicates the involvement of a peroxide. As shown in Figure 2, there is a clear indication that the photoactivity depends on reactive oxygen species (ROS). Both NaN3 and Trolox almost completely shut down photoreactivity of conjugate 1a (Figure 2, lanes 4 and 5, respectively). These data are in agreement with both singlet oxygen and peroxide as the ROS involved in the observed photoactivity. In the case of conjugate 1b, some attenuation in photoinduced DNA nicking is observed in the presence of NaN3 (Figure 2, compare lanes 6 and 7). In the presence of Trolox (Figure 2, lane 8), DNA nicking is attenuated but to a lesser extent. As the photoactivity of this conjugate (1b) is generally not substantial, it is difficult to conclude on the inhibitory effect of Trolox on the photoactivity of conjugate 1b.

DISCUSSION Photocleavage of a specific gene at a PDT-relevant wavelength may be a powerful tool for genetic modifications both in biotechnology and in gene therapy. Toward this goal, we have studied a series of TFO conjugates that differ in several characteristics: (1) the choice of sequence (i.e., wild type vs scrambled), (2) the intervening amino acid (i.e., Gly vs Trp), (3) the dye molecule (i.e., TO1 vs TO2), and (4) the ODN type (i.e., DNA vs DNA/LNA). Table 1 summarizes the photoactivity of the various conjugates. It is important to stress the contribution of the intervening amino acid. When glycine is placed between TFO and dye (conjugates 1b, 2b, and 3b), the only photoinduced DNA damage to the mdm2-plasmid is nicking of supercoiled DNA (form I) to its relaxed form (form II), an observation that is consistent with singlet oxygen reactivity. However, when L-tryptophan is introduced, photoactivation of the conjugates (1a, 2a, and 3a) in the presence of the mdm2 plasmid leads to its linearization; particularly for the LNA/DNA conjugate (conjugate 3a, Figure 4, lanes 2-5, and Table 1). This

620 Bioconjugate Chem., Vol. 21, No. 4, 2010

observation is highly relevant in the context of the cell as such DNA damage (i.e., double strand breaks) is less manageable by the cellular DNA repair mechanism (55). As Trolox inhibits this photoreaction (Figure 2, lanes 6) and inaccordancewithstudiesonthephotoactivityofpeptide-TO1(37,38), we conclude that a peroxide is formed on the indole ring of L-tryptophan leading to a radical reaction (most likely by 5′hydrogen abstraction (51)). It is this mechanism that distinguishes the photoactivity of L-tryptophan containing conjugates (1a, 2a, and 3a) from their glycine counterparts (1b, 2b, and 3b). The fact that linearization is more pronounced in the LNA/ DNA TFO could be related to the superior hybridization that is expected from such a TFO in comparison to DNA-only TFO’s. All conjugates with a scrambled TFO sequence (1c, 2c, and 3c) did not produce any detectable DNA damage upon photoactivation. This is indicative of a photoinduced reaction that is sequence-dependent. This is further supported by the photocleaved DNA product (lane 9, Figure 5) formed after irradiation of a synthetic 44-mer DNA duplex with the LNA TFO (conjugate 3a). Finally, when comparing both cyanine dyes (TO1 and TO2) we observe a similar photoactivity for all DNA conjugates (see Table 1). This is encouraging since, as opposed to TO1, TO2 has not been studied in the context of DNA damage. Thus, we have established the potential of utilizing cyanine dyes with longer absorption wavelengths for site-specific DNA photocleavage by TFO-Trp-Dye conjugates.

ACKNOWLEDGMENT We thank the Alex Grass Center for Drug Design and Synthesis of Novel Therapeutics for financial support. The PCMV-MDM2 plasmid was generously provided by Prof. Matthias Dobbelstein. We thank Shimon Steingart for technical assistance. J.K. and V.B. were supported by the Grant Agency of the Academy of Sciences of the Czech Republic (IAA400040803). Supporting Information Available: HPLC chromatograms of DNA and DNA/LNA conjugates and thermal denaturation curves of duplex (44-mer) and triplex (44-mer duplex and 14mer TFO). This material is available free of charge via the Internet at http://pubs.acs.org.

LITERATURE CITED (1) Besch, R., Giovannangeli, C., and Degitz, K. (2004) Triplexforming oligonucleotides - Sequence-specific DNA ligands as tools for gene inhibition and for modulation of DNA-associated functions. Curr. Drug Targets 5, 691–703. (2) Boerner, L. J. K., and Zaleski, J. M. (2005) Metal complexDNA interactions: from transcription inhibition to photoactivated cleavage. Curr. Opin. Chem. Biol. 9, 135–144. (3) Da Ros, T., Spalluto, G., Boutorine, A. S., Bensasson, R. V., and Prato, M. (2001) DNA-photocleavage agents. Curr. Pharm. Des. 7, 1781–1821. (4) Rogers, F. A., Lloyd, J. A., and Glazer, P. M. (2005) Triplexforming oligonucleotides as potential tools for modulation of gene expression. Curr. Med. Chem. Anti-Cancer Agents 5, 319– 326. (5) Thuong, N. T., and Helene, C. (1993) Sequence-specific recognition and modification of double-helical DNA by oligonucleotides. Angew. Chem., Int. Ed. 32, 666–690. (6) Barre, F. X., Giovannangeli, C., Helene, C., and Harel-Bellan, A. (1999) Covalent crosslinks introduced via a triple helixforming oligonucleotide coupled to psoralen are inefficiently repaired. Nucleic Acids Res. 27, 743–749.

Biton et al. (7) Cassidy, R. A., Kondo, N. S., and Miller, P. S. (2000) Triplex formation by psoralen-conjugated chimeric oligonucleoside methylphosphonates. Biochemistry 39, 8683–8691. (8) Christensen, L. A., Wang, H., Van Houten, B., and Vasquez, K. M. (2008) Efficient processing of TFO-directed psoralen DNA interstrand crosslinks by the UvrABC nuclease. Nucleic Acids Res. 36, 7136–7145. (9) Diviacco, S., Rapozzi, V., Xodo, L., Helene, C., Quadrifoglio, F., and Giovannangeli, C. (2001) Site-directed inhibition of DNA replication by triple helix formation. FEBS Lett. 15, 2660–2668. (10) Giovannangeli, C., Diviacco, S., Labrousse, V., Gryaznov, S., Charneau, P., and Helene, C. (1997) Accessibility of nuclear DNA to triplex-forming oligonucleotides: The integrated HIV-1 provirus as a target. Proc. Natl. Acad. Sci. U.S.A. 94, 79–84. (11) Higuchi, M., Kobori, A., Yamayoshi, A., and Murakami, A. (2009) Synthesis of antisense oligonucleotides containing 2 ′-Opsoralenylmethoxyalkyl adenosine for photodynamic regulation of point mutations in RNA. Bioorg. Med. Chem. 17, 475–483. (12) Kim, K. H., Nielsen, P. E., and Glazer, P. M. (2006) Sitespecific gene modification by PNAs conjugated to psoralen. Biochemistry 45, 314–323. (13) Kim, K. H., Nielsen, P. E., and Glazer, P. M. (2007) Sitedirected gene mutation at mixed sequence targets by psoralenconjugated pseudo-complementary peptide nucleic acids. Nucleic Acids Res. 35, 7604–7613. (14) Li, H., Broughton-Head, V. J., Peng, G. M., Powers, V. E. C., Ovens, M. J., Fox, K. R., and Brown, T. (2006) Triplex staples: DNA double-strand cross-linking at internal and terminal sites using psoralen-containing triplex-forming oligonucleotides. Bioconjugate Chem. 17, 1561–1567. (15) Majumdar, A., Khorlin, A., Dyatkina, N., Lin, F. L. M., Powell, J., Liu, J., Feiz, Z. Z., Khripine, Y., Watanabe, K. A., George, J., Glazer, P. M., and Seidman, M. M. (1998) Targeted gene knockout mediated by triple helix forming oligonucleotides. Nat. Genet. 20, 212–214. (16) Majumdar, A., Muniandy, P. A., Liu, J., Liu, J. I., Liu, S. T., Cuenoud, B., and Seidman, M. M. (2008) Targeted gene knock in and sequence modulation mediated by a psoralen-linked triplex-forming oligonucleotide. J. Biol. Chem. 283, 11244– 11252. (17) Musso, M., Wang, J. C., and VanDyke, M. W. (1996) In vivo persistence of DNA triple helices containing psoralen-conjugated oligodeoxyribonucleotides. Nucleic Acids Res. 24, 4924–4932. (18) Oh, D. H., Suzara, V., and Krishnan, R. (2008) Modulation of psoralen DNA crosslinking kinetics associated with a triplexforming oligonucleotide. Photochem. Photobiol. 84, 727–733. (19) Shahid, K. A., Majumdar, A., Alam, R., Liu, S. T., Kuan, J. Y., Sui, X. F., Cuenoud, B., Glazer, P. M., Miller, P. S., and Seidman, M. M. (2006) Targeted cross-linking of the human betaglobin gene in living cells mediated by a triple helix forming oligonucleotide. Biochemistry 45, 1970–1978. (20) Takasugi, M., Guendouz, A., Chassignol, M., Decout, J. L., Lhomme, J., Thuong, N. T., and Helene, C. (1991) Sequencespecific photoinduced cross-linking of the 2 strands of doublehelical DNA by a psoralen covalently linked to a triple helixforming oligonucleotide. Proc. Natl. Acad. Sci. U.S.A. 88, 5602– 5606. (21) Vasquez, K. M., Wang, G., Havre, P. A., and Glazer, P. M. (1999) Chromosomal mutations induced by triplex-forming oligonucleotides in mammalian cells. Nucleic Acids Res. 27, 1176–1181. (22) Vasquez, K. M., Wensel, T. G., Hogan, M. E., and Wilson, J. H. (1996) High-efficiency triple-helix-mediated photo-crosslinking at a targeted site within a selectable mammalian gene. Biochemistry 35, 10712–10719. (23) Yamayoshi, A., Kato, K., Suga, S., Ichinoe, A., Arima, T., Matsuda, T., Kato, H., Murakami, A., and Wake, N. (2007) Specific apoptosis induction in human papillomavirus-positive cervical carcinoma cells by photodynamic antisense regulation. Oligonucleotides 17, 66–79.

DNA Photocleavage by DNA and DNA-LNA Conjugates (24) Ye, Z. Y., Guntaka, R. V., and Mahato, R. I. (2007) Sequencespecific triple helix formation with genomic DNA. Biochemistry 46, 11240–11252. (25) Crean, C. W., Kavanagh, Y. T., O’Keeffe, C. M., Lawler, M. P., Stevenson, C., Davies, R. J. H., Boyle, P. H., and Kelly, J. M. (2002) Targeting of photooxidative damage on singlestranded DNA representing the bcr-abl chimeric gene using oligonucleotide-conjugates containing [Ru(phen)(3)](2+)-like photosensitiser groups. Photochem. Photobiol. Sci. 1, 1024–1033. (26) Doan, T. L., Perrouault, L., Praseuth, D., Habhoub, N., Decout, J. L., Thuong, N. T., Lhomme, J., and Helene, C. (1987) Sequence-specific recognition, photo-cross-linking and cleavage of the DNA double helix by an oligo-[alpha]-thymidylate covalently linked to an azidoproflavine derivative. Nucleic Acids Res. 15, 7749–7760. (27) Inoue, T., Sugiura, Y., Saitoh, J., Ishiguro, T., and Otsuka, M. (1999) Fluorescence property of oxazole yellow-linked oligonucleotide. Triple helix formation and photocleavage of double-stranded DNA in the presence of spermine. Bioorg. Med. Chem. 7, 1207–1211. (28) Le Doan, T., Praseuth, D., Perrouault, L., Chassignol, M., Thuong, N. T., and Helene, C. (1990) Sequence-targeted photochemical modifications of nucleic acids by complementary oligonucleotides covalently linked to porphyrins. Bioconjugate Chem. 1, 108–113. (29) Le Gac, S., Rickling, S., Gerbaux, P., Defrancq, E., Moucheron, C., and, and Kirsch-De Mesmaeker, A. (2009) A photoreactive ruthenium(II) complex tethered to a guanine-containing oligonucleotide: A biomolecular tool that behaves as a ”seppuku molecule. Angew. Chem., Int. Ed. 48, 1122–1125. (30) Lentzen, O., Constant, J. F., Defrancq, E., Prevost, M., Schumm, S., Moucheron, C., Dumy, P., and Kirsch-De Mesmaeker, A. (2003) Photocrosslinking in ruthenium-labelled duplex oligonucleotides. ChemBioChem 4, 195–202. (31) Ortmans, I., Content, S., Boutonnet, N., Kirsch-De Mesmaeker, A., Bannwarth, W., Constant, J. F., Defrancq, E., and, and Lhomme, J. (1999) Ru-labeled oligonucleotides for photoinduced reactions on targeted DNA guanines. Chem.sEur. J. 5, 2712– 2721. (32) Perrouault, L., Asseline, U., Rivalle, C., Thuong, N. T., Bisagni, E., Giovannangeli, C., Ledoan, T., and Helene, C. (1990) Sequence-specific artificial photo-induced endonucleases based on triple helix-forming oligonucleotides. Nature 344, 358–360. (33) Boutorine, A. S., Brault, D., Takasugi, M., Delgado, O., and Helene, C. (1996) Chlorin-oligonucleotide conjugates: Synthesis, properties, and red light-induced photochemical sequence-specific DNA cleavage in duplexes and triplexes. J. Am. Chem. Soc. 118, 9469–9476. (34) Magda, D., Wright, M., Miller, R. A., Sessler, J. L., and Sansom, P. I. (1995) Sequence-specific photocleavage of DNA by an expanded porphyrin with irradiation above 700 nm. J. Am. Chem. Soc. 117, 3629–3630. (35) Sessler, J. L., Sansom, P. I., Kral, V., Oconnor, D., and Iverson, B. L. (1996) Sapphyrin-oligonucleotide conjugates. Novel sequence-specific DNA photomodifying agents with increased binding affinity. J. Am. Chem. Soc. 118, 12322–12330. (36) Nygren, J., Svanvik, N., and Kubista, M. (1998) The interactions between the fluorescent dye thiazole orange and DNA. Biopolymers 46, 39–51. (37) Mahon, K. P., Ortiz-Meoz, R. F., Prestwich, E. G., and Kelley, S. O. (2003) Photosensitized DNA cleavage promoted by amino acids. Chem. Commun. 1956–1957. (38) Mahon, K. P., Roy, M. D., Carreon, J. R., Prestwich, E. G., Rouge, J. L., Shin, S., and Kelley, S. O. (2006) Tunable DNA

Bioconjugate Chem., Vol. 21, No. 4, 2010 621 cleavage by intercalating peptidoconjugates. ChemBioChem 7, 766–773. (39) Carreon, J. R., Stewart, K. M., Mahon, K. P., Shin, S., and Kelley, S. O. (2007) Cyanine dye conjugates as probes for live cell imaging. Bioorg. Med. Chem. Lett. 17, 5182–5185. (40) Brunet, E., Alberti, P., Perrouault, L. C., Babu, R., Wengel, J., and Giovannangeli, C. (2005) Exploring cellular activity of locked nucleic acid-modified triplex-forming oligonucleotides and defining its molecular basis. J. Biol. Chem. 280, 20076–20085. (41) Kumar, N., Nielsen, K. E., Maiti, S., and Petersen, M. (2006) Triplex formation with alpha-L-LNA (alpha-L-ribo-configured locked nucleic acid). J. Am. Chem. Soc. 128, 14–15. (42) Kumar, N., Patowary, A., Sivasubbu, S., Petersen, M., and Maiti, S. (2008) Silencing c-MYC expression by targeting quadruplex in P1 promoter using locked nucleic acid trap. Biochemistry 47, 13179–13188. (43) Petersen, M., and Wengel, J. (2003) LNA: a versatile tool for therapeutics and genomics. Trends Biotechnol. 21, 74–81. (44) Rahman, S. M. A., Seki, S., Obika, S., Haitani, S., Miyashita, K., and Imanishi, T. (2007) Highly stable pyrimidine-motif triplex formation at physiological pH values by a bridged nucleic acid analogue. Angew. Chem., Int. Ed. 46, 4306–4309. (45) Sun, B. W., Babu, B. R., Sorensen, M. D., Zakrzewska, K., Wengel, J., and Sun, J. S. (2004) Sequence and pH effects of LNA-containing triple helix-forming oligonucleotides: Physical chemistry, biochemistry, and modeling studies. Biochemistry 43, 4160–4169. (46) Torigoe, H., Hari, Y., Sekiguchi, M., Obika, S., and Imanishi, T. (2001) 2 ′-O,4 ′-C-methylene bridged nucleic acid modification promotes pyrimidine motif tripler DNA formation at physiological pH - Thermodynamic and kinetic studies. J. Biol. Chem. 276, 2354–2360. (47) Bianco, R., Ciardiello, F., and Tortora, G. (2005) Chemosensitization by antisense oligonucleotides targeting MDM2. Curr. Cancer Drug Targets 5, 51–56. (48) Rayburn, E., Zhang, R. W., He, J., and Wang, H. (2005) MDM2 and human malignancies: Expression, clinical pathology, prognostic markers, and implications for chemotherapy. Curr. Cancer Drug Targets 5, 27–41. (49) Toledo, F., and Wahl, G. M. (2006) Regulating the p53 pathway: in vitro hypotheses, in vivo veritas. Nat. ReV. Cancer 6, 909–923. (50) Stern, M., Warshawsky, A., and Fridkin, M. (1979) Facile thiolytic removal of the ortho-nitrophenylsulfenyl amino-protecting group. Int. J. Pept. Protein Res. 13, 315–319. (51) Prestwich, E. G., Roy, M. D., Rego, J., and Kelley, S. O. (2005) Oxidative DNA strand scission induced by peptides. Chem. Biol. 12, 695–701. (52) Sau, S. P., Kumar, P., Anderson, B. A., Østergaard, M. E., Deoblad, L., Paszczynski, A., Sharma, P. K., and Hrdlicka, P. J. (2009) Optimized DNA-targeting using triplex forming C5alkynyl functionalized LNA. Chem. Commun. 6756–6758. (53) Hasty, N., Merkel, P. B., Radlick, P., and Kearns, D. R. (1972) Role of azide in singlet oxygen reactions - reaction of azide with singlet oxygen. Tetrahedron Lett. 49–52. (54) Sarker, A. H., Watanabe, S., Seki, S., Akiyama, T., and Okada, S. (1995) Oxygen radical-induced single-strand DNA breaks and repair of the damage in a cell-free system. Mutation Res. DNA Repair 337, 85–95. (55) Pardo, B., Gomez-Gonzalez, B., and Aguilera, A. (2009) DNA repair in mammalian cells. CMLS 66, 1039–1056. BC900372H