Caged Single and Double Strand Breaks - Bioconjugate Chemistry

94

Bioconjugate Chem. 2000, 11, 94−103

Caged Single and Double Strand Breaks Phillip Ordoukhanian and John-Stephen Taylor* Department of Chemistry, Washington University, St. Louis, Missouri 63130. Received July 23, 1999; Revised Manuscript Received October 8, 1999

Ionizing radiation and radiomimetic drugs such as bleomycin, calichieamycin, neocarzinostatin chromophore, and other synthetic agents can produce both single and double strand breaks in DNA. The ability to study the structure-activity relationships of single and double-strand break repair, lethality, and mutagenesis in vivo is complicated by the numerous types and sites of DNA cleavage products that can be induced by such agents. The ability to “cage” such breaks in DNA might help to further such studies and additionally afford a mechanism for activating and deactivating nucleic acid based drugs and probes. The major type of single strand break induced by ionizing radiation is a 3′and 5′-phosphate terminated single nucleotide gap. Previously, a caged strand break of this type had been developed that was designed to produce the 5′-phosphate directly upon irradiation with 366 nm light, and the 3′-phosphate by a subsequent β-elimination reaction [Ordoukhanian, P., and Taylor, J.-S. (1995) J. Am. Chem. Soc. 117, 9570]. Unfortunately, the release of the 3′-phosphate group was quite slow at pH 7. To circumvent this problem, a second caged strand break has been developed that produces the 3′-phosphate directly upon irradiation, and the 5′-phosphate by a subsequent β-elimination reaction. When this caged strand break was used in tandem with the previous caged strand break, 5′and 3′-phosphate terminated gaps could be directly produced by irradiation with 366 nm light. These caged single strand breaks were also incorporated in tandem into hairpin substrates to demonstrate that they could be used to cage double strand breaks. These caged single strand breaks should be generally useful for generating site-specific DNA single and double strand breaks and gaps, using wavelengths and doses of light that are nondetrimental to biological systems. Because the position of the single strand break can be varied, it should now be possible to examine the effect of the sequence context and cleavage pattern of single and double strand breaks on the lethality and mutagenicity of this important class of DNA damage.

INTRODUCTION

Much of the lethality of ionizing radiation and radiomimetic agents has been attributed to double strand breaks (1-5). Little is known, however, about the structure-activity relationships in this process because of a lack of methods for producing site-specific double strand breaks with specific cleavage patterns and termini. In most studies of double strand break repair, the double strand breaks are produced by restriction enzyme cleavage (6), which results in a directly religatable break (5′phosphate and 3′-hydroxyl terminated nick) with a limited pattern of cleavage. Ionizing radiation, on the other hand, often produces phosphate-terminated single nucleotide gaps that result from C4′ radical assisted solvolysis of the 5′ and 3′ phosphates (Figure 1), which cannot be directly religated (1, 7). The same phosphateterminated gap can also be produced by AP lyase activity on an abasic site formed by glycosylases which process ionizing radiation damaged bases in base excision repair (BER) pathway (8). Because double strand breaks are inherently unstable, a method for triggering the formation of site-specific double strand breaks of varying termini and sequence context would facilitate their construction and subsequent in vitro or in vivo activation. Photorelease of “caged” substrates, drugs, and enzymes has been increasingly used as a means of triggering biological events (see articles in ref 9), since the first * To whom correspondence should be addressed. Phone: (314) 935-6721. Fax: (314) 935-4481. E-mail: taylor@ wuchem.wustl.edu.

reports of caged ATP in the late 1970s (10, 11). Molecules are generally caged by covalent attachment to a photochemically cleavable auxiliary that renders the molecule inactive as a substrate, drug, or enzyme, and allows it to diffuse into the system of interest. The photochemical auxiliary is chosen so that it can be released from the molecule of interest by irradiation under wavelengths and fluences that are not injurious to the system under study. The rate of release can be controlled by the amplitude or the wavelength of light used and structural modifications that increase the quantum yield of cleavage. Caged molecules are typically used for kinetic investigations in biological systems where rapid mixing techniques are impractical, such as, inside cells or tissues. Native and damaged DNA are the substrates for a wide variety of enzymes, most notably those involved in replication, transcription, and repair, and caged substrates could play an important role in studying the mechanisms of such processes. RNA molecules also interact with a wide variety of enzymes, some have their own enzymatic activity, and their study could also be aided by the availability of caged RNA substrates. Though notable advances have been made in caging cofactors, drugs, and proteins, there are only a few examples of site-specifically caged polynucleotide substrates. For DNA, a caged abasic site (12), and deoxyribonolactone (13) have been reported, and for RNA, a caged ribozyme has been reported (14). A number of years ago, we reported the synthesis of a photocleavable DNA building block based on a cis-syn thymine dimer lacking an internucleotide phosphate (15).

10.1021/bc9900993 CCC: $19.00 © 2000 American Chemical Society Published on Web 12/23/1999

Caged Single and Double Strand Breaks

Bioconjugate Chem., Vol. 11, No. 1, 2000 95

Figure 1. (a) Direct and indirect formation of a single strand break consisting of a one nucleotide gap terminating in a 5′-and 3′-phosphate by ionizing radiation. In the direct pathway, hydrogen abstraction of the H4′ hydrogen by hydroxyl radical causes the stepwise solvolysis of both 5′- and 3′-phosphates. In the indirect pathway, the base portion of a nucleotide becomes damaged which results in spontaneous or enzymatic glycosidic bond hydrolysis to yield an abasic site which releases the phosphates by β-elimination reactions catalyzed by an enzymatic AP lyase activity. (b) Examples of strand breaks that can be produced directly or indirectly by exposure of double strand DNA to ionizing radiation in vivo.

Strand cleavage could be induced by either direct photoreversion of the thymine dimer by irradiation with 254 nm light or by irradiation with 400 nm light in the presence of photolyase. This building block is not generally useful, however, because direct photoreversal with 254 nm light also damages DNA, and because the enzymatic photoreversal reaction was not very efficient and is limited to dithymine sequences in DNA. Subsequently, a C4′-phenyl selenide adenosine derivative was shown to give a phosphate-terminated one nucleotide gap upon photolysis at >320 nm under anaerobic and nonreducing conditions (7), but unfortunately, under aerobic or reducing conditions, a number of other side products were also obtained (16). To circumvent the drawbacks of the previously reported building blocks, we designed an o-nitrobenzyl-based photocleavable building block 1 with the same backbone as in DNA (Figure 2a) (17).1 This building block was designed to introduce a caged strand break X that would produce a phosphate-terminated gap

by a two step reaction initiated by irradiation with 360 nm light. Photolysis of 2 with 366 nm light was expected to directly release the DNA fragment 4 terminating in a 5′-phosphate and another terminating in the 3′-β-ketophosphate derivative 3. The latter product was then expected to release the DNA fragment 5 terminating in a 3′-phosphate by a β-elimination reaction (Figure 2a). When an oligonucleotide containing this building block was irradiated at 366 nm, the DNA fragment 4, terminating in a 5′-phosphate, was immediately released, but the release of the DNA fragment 5 terminating in a 3′phosphate did not occur spontaneously at pH 7 and required heating with piperidine. Herein, we report the design, synthesis, and use of a second o-nitrobenzyl building block 11 for introducing a 1 A similar building block based on 1-(o-nitrophenyl)-1,2ethanediol has been described by M. S. Urdea and T. Horn in U.S. Patent 5,258,506.

96 Bioconjugate Chem., Vol. 11, No. 1, 2000

Ordoukhanian and Taylor

Figure 2. (a) Scheme for the synthesis of building block 1 used to incorporate the caged single strand break X into oligodeoxynucleotides, and the expected photolysis products of an oligodeoxynucleotide containing X. (b) An autoradiogram of a denaturing polyacrylamide electrophoresis gel of the products from exposing 5′-end labeled oligodeoxynucleotide 2 at pH 7.0, 8.0, and 9.3 (Tris-HCl) to 366 nm light from a hand held UV-lamp for 15 min at room temperature, followed by incubation at the times and temperatures shown prior to gel electrophoresis at pH 7.0. The fastest moving band comigrates with product 5 that was produced by a Maxam-Gilbert G reaction on the oligodeoxynucleotide corresponding to 2 in which X was replaced with pdG (Lane G). Only the second slowest moving band appeared to be readily converted to the product and was assigned as intermediate 3. The slowest migrating band U, which is the major product of irradiation at pH 7, was not readily converted to 5.

Figure 3. (a) Synthesis of the building block 11 used to incorporate a caged single strand break Y into oligodeoxynucleotides, and the expected photolysis products of oligodeoxynucleotides containing Y. On the right are autoradiograms of denaturing polyacrylamide electrophoresis gels of the products of exposing (b) 5′-end and (c) 3′-end labeled oligodeoxynucleotides 12a and 12b at pH 7.0 and room temperature to 366 nm light from a hand held UV-lamp for the indicated time in min. The products of the 16 min irradiation were also subjected to 1 M piperidine at 90 °C for 30 min (lane P). The fastest moving band in panel b comigrates with the product 5 that was produced by a Maxam-Gilbert G reaction on the oligodeoxynucleotide corresponding to 12a in which Y was replaced with pdG (lane G).

caged strand break Y that leads directly to a 3′-phosphate terminated DNA fragment upon irradiation (Figure 3a). We also show that when the caged strand break Y is linked in a tandem fashion with the caged strand break X, phosphate-terminated gaps can be produced directly by photolysis (Figure 4). In addition, we demonstrate that these building blocks can be used to efficiently induce double strand breaks in DNA.

MATERIALS AND METHODS

All reactions were conducted under anhydrous conditions and an argon atmosphere. Methylene chloride and THF2 were dried by distillation from the appropriate drying reagents. The 1 M solution of TBAF in THF was obtained from Aldrich. 1H NMR and 13C NMR spectra were acquired on a Varian Gemini-300 spectrometer, 31P NMR spectra were acquired on a Varian XL-300 spec-



Figure 4. Scheme illustrating how two caged single strand breaks X and Y can be used to produce 3′- and 5′-phosphate terminated gaps directly upon photolysis.

trometer. Low- and high-resolution mass spectra were obtained from the Washington University Resource for Biomedical and Bio-organic Mass Spectrometry, Department of Chemistry, Washington University. Oligodeoxynucleotides were synthesized on an ABI 380B synthesizer by standard β-cyanoethyl phosphoramidite chemistry and purified on a preparative, 2 mm thick 165 mm long, 7 M urea, 1:19 cross-linked, 19% acrylamide gel at 300 V. The coupling yield for the modified building block 11 was >98%, with a 2 h coupling time. Photosensitive oligodeoxynucleotides were visualized on the preparative gel by doping with 5′-[32P]-end labeled oligonucleotide and autoradiography. Nonphotosensitive oligodeoxynucleo2 Abbreviations: BER, base excision repair; ddATP, dideoxyadenosinetriphosphate; DBU, diazabicycloundecane; DMT, 4, 4′dimethoxytrityl; DTT, dithiothreitol; NER, nucleotide excision repair; TBAF, tetrabutylammonium fluoride; TBDMS, tertbutyldimethylsilyl; TEA, triethylamine; THF, tetrahydrofuran; TMP, trimethyl phosphate; TMS, tetramethylsilane; UVA, 320400 nm.

tides were visualized by shadowing with 254 nm light against a fluorescent TLC plate. All oligodeoxynucleotide concentrations were determined by UV absorbance measurements (18). Oligodeoxynucleotides were 5′-end labeled with [γ-32 P]ATP (Amersham) and T4 polynucleotide kinase (New England Biolabs) in 70 mM Tris-HCl, pH 7.6, 10 mM MgCl2, and 5 mM DTT according to standard procedures (19). Oligodeoxynucleotides were 3′end labeled with [R-32P]ddATP (Amersham) and terminal deoxynucleotide transferase in 100 mM sodium cacodylate, pH 7.2, 0.2 mM 2-mercaptoethanol, and 2 mM CoCl2‚6H2O. Alternatively, the photocleavable oligodeoxynucleotides were 3′-end labeled by annealing to the complementary 10-mer, d(TTCGACGTAT), and gap filling with [R-32P]dATP and 1 unit of DNA polymerase I large (Klenow) fragment (New England Biolabs) in 10 mM Tris-HCl, pH 7.5, and 5 mM MgCl2. Photolyses were conducted in the lid of a polyethylene microcentrifuge tube with a model UVGL-25 (VWR) hand-held light with an intensity of approximately 0.75 mW/cm2 (0.21 mW).


Photolysis products were analyzed by electrophoresis on 0.4 mm thick, 31 cm wide, 38.5 cm long, 7 M urea, 1:19 cross-linked, 19% or 15% acrylamide gels. Radiolabeled products were visualized by autoradiography with Kodak XAR-5 film at -70 °C or, for quantitative purposes, with a Molecular Dynamics Phosphoimager. 1-o-Nitrophenyl-1-O-tert-butyldimethylsilyl-1,3propanediol (8). To a stirred solution of 1 g (5.06 mmol) of the diol 7 (17) in 15 mL methylene chloride was added 517 mg (7.6 mmol) of imidazole. After the imidazole had all dissolved, 1.15 g (7.6 mmol) of TBDMS chloride was added and allowed to stir for an additional 2 h. A saturated Na2CO3 solution was added to quench the reaction, and the organic layer was separated. The aqueous layer was further extracted with methylene chloride, the organic layers were combined and dried with MgSO4, and the solvent was removed under reduced pressure. The resulting residue was chromatographed on a silica gel column (8:2 hexane/ethyl acetate) to give the desired compound 8 (1.39 g, 4.5 mmol) in 88% yield: Rf ) 0.54 (8:2 hexane/ethyl acetate); 1H NMR (300 MHz, CD3COCD3, referenced to TMS): δ 7.90 (d, J ) 8 Hz, 1H, ArNO2), 7.86 (d, J ) 8 Hz, 1H, ArNO2), 7.71 (t, J ) 8 Hz, 1H, ArNO2), 7.49 (td, J ) 8, 1 Hz, 1H, ArNO2), 5.405.33 (m, 1H), 4.61 (d, J ) 4 Hz, 1H), 3.98-3.79 (m, 2H), 2.06-1.92 (m, 1H), 1.89-1.75 (m, 1H), 0.91 (s, 9H), 0.08 (s, 6H); 13C NMR (75 MHz, CD3COCD3, referenced to TMS): δ 148.8, 141.7, 133.8, 129.0, 128.6, 124.5, 66.7, 66.6, 61.0, 42.3, 26.2, 18.7, -5.3, -5.2; IR (neat) 3441, 2946, 2857, 1654, 1647, 1636, 1594, 1566, 1560, 1491, 1455, 1437, 1399, 1376, 1313, 1282 cm-1; FAB HRMS m/z C15 H26 O4 N Si, [M + H]+ calcd, 312.1631; found, 312.1633. 1-o-Nitrophenyl-1-O-dimethoxytrityl -3-O-tert-butyldimethylsiyl-1,3-propanediol (9). To a stirred solution of 0.51 g (1.6 mmol) of 8 in 10 mL methylene chloride was added 0.34 mL (2.3 mmol) of DBU, followed by 0.72 g (2.1 mmol) of DMT chloride. The reaction was allowed to stir for 2 days at room temperature, after which a saturated Na2CO3 solution was added and the organic layer was separated. The aqueous layer was further extracted with ethyl acetate, the organic layers were combined and dried with MgSO4, and the solvent was removed under reduced pressure. The resulting residue was chromatographed on a silica gel column (8.9:0.9:0.2 hexane/ethyl acetate/TEA) to give 9 (0.91 g, 1.5 mmol) in 90% yield: Rf ) 0.57 (8.9:0.9:0.2 hexane/ethyl acetate/ TEA); 1H NMR (300 MHz, CD2Cl2, referenced to TMS) δ 7.58 (d, J ) 8 Hz, 2H, Ar), 7.46 (d, J ) 8 Hz, 2H, Ar), 7.37-7.07 (m, 9 H, Ar), 6.66 (d, J ) 9 Hz, 2 H, Ar), 6.60 (d, J ) 9 Hz, 2 H, Ar), 5.43-5.26 (m, 1H), 3.74-3.64 (m, 8H), 2.18 (quartet, J ) 7 Hz, 2 H), 0.82 (s, 9H), -0.03 (s, 6H); 13C NMR (75 MHz, CD2Cl2, referenced to TMS) δ 158.9, 158.8, 146.9, 146.2, 140.7, 136.5, 136.2, 132.7, 130.7, 130.6, 130.6, 127.0, 124.0, 113.3, 113.1, 87.6, 69.6, 59.7, 55.5, 41.6, 26.0, 18.4, -5.2; IR (neat) 3853, 3744, 3059, 3035, 3000, 2954, 2930, 2898, 2884, 2855, 2837, 1608, 1525, 1509, 1464, 1445, 1359, 1340, 1302, 1252, 1177, 1092, 1036, 1009 cm-1; FAB HRMS m/z C36 H43 O6 N Si Na [M + Na]+ calcd, 636.2757; found, 636.2758. 1-o-Nitrophenyl-1-O-dimethoxytrityl-1,3-propanediol (10). To a stirred solution of 0.9 g (1.5 mmol) of 9 in 8 mL of THF was added 2.3 mL (2.3 mmol) of a 1 M solution of TBAF in THF. The reaction was allowed to stir for 2 h, after which the solvent was removed under reduced pressure. The resulting residue was chromatographed on a silica gel column (7.9:1.9:0.2 hexane/ethyl acetate/TEA) to give 10 (0.56 g, 1.1 mmol) as a white foam in 77% yield: Rf ) 0.17 (7.9:1.9:0.2 hexane/ethyl acetate/


TEA); 1H NMR (300 MHz, CD3COCD3, referenced to TMS) δ 7.74-7.61 (m, 2H), 7.54-7.48 (m, 2H, Ar), 7.44 (td, J ) 8, 1 Hz, 1H, Ar), 7.31 (dd, J ) 9, 2 Hz, 2H, Ar), 7.27-7.11 (m, 6H), 6.72 (dd, J ) 9, 2 Hz, 2H), 6.65 (dd, J ) 9, 2 Hz, 2H), 5.39 (m, 1H), 3.72 (overlapping s, 6H, ArOCH3), 3.69-3.53 (m, 2H), 3.45 (br s, 1H), 2.35-2.10 (m, 2H); 13C NMR (75 MHz, CD3COCD3, referenced to TMS) δ 184.7, 159.4, 159.3, 147.4, 146.5, 140.5, 136.7, 136.4, 133.1, 130.9, 130.9, 130.8, 128.6, 128.3, 127.6, 127.3, 124.3, 113.6, 113.5, 88.1, 70.1, 70.0, 58.5, 55.2, 41.6; IR (neat) 3416, 3406, 3241, 3058, 3036, 3001, 2956, 2934, 2908, 2837, 1701, 1607, 1577, 1522, 1509, 1464, 1445, 1357, 1340, 1302, 1251, 1225, 1177, 1153 cm-1; FAB HRMS m/z C30 H29 O6 N Na [M + Na]+ calcd, 522.1892; found, 522.1893. Photocleavable DNA Building Block (11). To a stirred solution of 0.18 g (0.37 mmol) of 10 in 5 mL methylene chloride were added 0.2 mL TEA (1.4 mmol) and 0.19 mL (0.85 mmol) of N,N-diisopropyl-(2-cyanoethyl)-phosphonamidic chloride. The reaction was allowed to stir for and additional 30 min, after which a saturated Na2CO3 solution was added. The organic layer was separated, and the aqueous layer further extracted with ethyl acetate. The organic layers were combined and dried with MgSO4. The solvent was removed under reduced pressure, and the resulting residue chromatographed on a silica gel column (7.9:1.9:0.2 hexane/ethyl acetate/TEA) to give 11 (0.24 g, 0.34 mmol) as a white foam in 93% yield: Rf ) 0.37 (7.9:1.9:0.2 hexane/ethyl acetate/TEA); 31P NMR (121.5 MHz, CD2Cl2, referenced to TMP) δ 144.3, 144.3; 1H NMR (300 MHz, CD2Cl2, referenced to TMS) δ 7.68-7.57 (m, 2H, Ar), 7.49-7.42 (m, 2H, Ar), 7.34 (t, J ) 8 Hz, 1H, Ar), 7.30-7.09 (m, 8H), 6.66 (dd, J ) 9, 2 Hz, 2H), 6.60 (dd, J ) 9, 2 Hz, 2H), 5.32 (br, s, 1H), 3.84-3.62 (m, 10 H), 3.61-3.43 (m, 2H), 2.53 (t, J ) 3 Hz, 2H), 2.32 (m, 2H), 1.29-1.02 (m, 13 H); 13C NMR (75 MHz, CD2Cl2, referenced to TMS) δ 160.5, 160.4, 148.5, 147.7, 141.9, 137.9, 137.6, 134.4, 132.2, 132.1, 129.7, 129.7, 129.6, 128.7, 128.6, 125.6, 119.8, 114.9, 114.8, 89.3, 71.1, 71.0, 61.5, 61.5, 61.3, 61.3, 60.4, 60.4, 60.2, 60.1, 57.1, 44.9, 44.7, 41.3, 41.2, 41.1, 26.3, 26.2, 26.2, 26.1, 22.3, 22.2; IR (neat) 2965, 2933, 2874, 1680, 1673, 1666, 1660, 1650, 1642, 1631, 1608, 1590, 1563, 1552, 1525, 1509, 1480, 1451, 1426, 1406, 1389, 1364, 1350, 1322, 1275, 1252, 1231, 1210, 1195, 1179, 1163, 1154, 1139, 1121, 1072, 1012, 996 cm-1. Photolysis of d(AATTGCATAXATACGTCGA) (2), at pH 7.0, 8.0, and 9.3. Three 30 µL solutions, ∼0.27 mM in oligodeoxynucleotide 2 (17) and 100 mM in TrisHCl at either pH 7.0, 8.0, or 9.3, were photolyzed with 366 nm light for 15 min at 0 °C. Each solution was divided into two portions, one of which was incubated at room temperature and the other at 37 ˚C. Aliquots of about 3 µL were taken immediately after photolysis and at 0.5, 1.0, and 2.0 h of incubation. All aliquots were stored at -20 °C until they could be loaded on a pH 7 denaturing 19% polyacrylamide electrophoresis gel. The products were also compared to the products of a MaxamGilbert G reaction (20) that was carried out on an otherwise identical sequence in which X was replaced with pdG. Photolysis of the 5′-End Labeled d(AATTGCATAYATACGTCGA) (12a) and 3′-End Labeled d(AATTGCATAYATACGTCGAdA) (12b). Two 50 µL solutions of either oligodeoxynucleotide 12a (24 nM) or oligodeoxynucleotide 12b (30 nM) in 100 mM in Tris-HCl (pH 7.0) were photolyzed with 366 nm light at room temperature. Aliquots of about 4 mL were taken following 0, 1, 2, 4, 8, and 16 min of irradiation and electrophoresed


on a 19% polyacrylamide gel. An additional final aliquot was taken and diluted up to 50 mL in a 1 M piperidine solution and heated to 90 °C for 30 min. The products were also compared to the products of a Maxam-Gilbert G reaction (20) that was carried out on an otherwise identical sequence in which Y was replaced with pdG. Photolysis of 5′-End Labeled d(AATTGCATAYAnXATACGTCGA) (14a, n ) 0, 1, or 4), and 3′-End Labeled d(AATTGCATAYXATACGTCGAA) (14b, n ) 0). Two 60 µL solutions of either oligonucleotide 14a (20 nM) or oligonucleotide 14b (20 nM) in 100 mM in Tris-HCl (pH 7.0) were photolyzed at 366 nm and room temperature. Aliquots of about 4 µL were taken at 0, 1, 2, 4, 8, and 16 min of irradiation and electrophoresed on a 19% polyacrylamide gel. An additional final aliquot was taken and diluted up to 50 mL in a 1 M piperidine solution and heated to 90 °C for 30 min. The products were also compared to the products of a Maxam-Gilbert G reaction (20) that was carried out on an otherwise identical sequence to 14a (n ) 0) in which X and Y were replaced by a single pdG. Preparation of Internally Labeled Hairpins 18 and 19 Containing Opposed Photocleavable Building Blocks X and Y. Two 40 µL solutions of d(TATCGAATTGTTXGGTACCCTTTCCCAAAAAGGGAAAGG) (3.6 mM) and either the 5′-[32P]-end labeled d(GTACCTYACAATTCG) (0.25 mM) or 5′-[32P]-end labeled d(GTACCTAACYATTCG) (0.25 mM), were combined with 6 Weiss units of T4 DNA Ligase (Promega) and 1.25 mM ATP, 50 mM Tris-HCl (pH 7.8), 10 mM MgCl2, 10 mM DTT, and 50 µg/mL BSA. The mixture was incubated for ∼40 h at 0 °C to give the ligated 54 mer hairpins 18 and 19, respectively. Photolysis of the Internally Labeled Hairpins 18 and 19. A 30 µL solution of either oligodeoxynucleotide 18 or 19 (33 nM) in 100 mM Tris-HCl (pH 7.0) and 25 mM NaCl, was annealed by heating to 95 °C for 3 min and allowed to cool to room temperature over ∼2 h. The annealed hairpin oligodeoxynucleotides were photolyzed at 366 nm and room temperature, and 3 µL aliquots taken at 0, 1, 2, 4, 8, and 16 min of irradiation and electrophoresed on a 19% polyacrylamide gel. Photolysis under denaturing conditions was carried out in the same way, except that the solution contained 0.15 N NaOH in place of the Tris-HCl buffer. RESULTS AND DISCUSSION

Before embarking on the synthesis of a new building block to directly release a DNA fragment terminating in a 3′-phosphate, we first reexamined the photochemistry of the original caged strand break X at different pHs, with the expectation that the β-elimination reaction should occur more rapidly at higher pH (Figure 2a). When 5′-end labeled oligodeoxynucleotide 2 was photolyzed at pH 7, three intermediate bands appeared, the slowest moving of which was the major product (Figure 2b). As we had found previously, the major product, which was originally assigned to structure 3, was stable to incubation at pH 7 for 2 h at 37 °C. When oligodeoxynucleotide 2 was photolyzed at pH 8, however, the major product was now the second most slowly moving band. This second band converted to the 3′-phosphate terminated product 5 slowly at room temperature, and more rapidly at 37 °C. Product 5 was produced independently by conducting a Maxam Gilbert G reaction (20) on an oligodeoxynucleotide identical in sequence to 2 except that X was replaced by G. When irradiated at pH 9, the 3′-phosphate-terminated product was the major product,


even at short incubation times. These results suggest that the major product at pH 7 is not the desired β-ketophosphate intermediate 3. Furthermore, the results suggest that intermediate 3 is only produced at pH 8 or above, and that it only β-eliminates rapidly at pH 9. We do not know the structure of the pH 7 product at this time. To directly photogenerate a 3′-phosphate-terminated fragment, we designed a second building block 11 (Figure 3a) that is based on the synthetic route used to prepare the original photocleavable building block 1 (Figure 2a) (17). The building block 11 was synthesized in four steps, in an overall yield of ∼60%, starting from the same intermediate 7 used in the synthesis of 1. The building block 11 was used to incorporate the caged strand break Y into a 19-mer oligodeoxynucleotide by automated synthesis, which was then 5′-end labeled with [γ-32P]dATP and polynucleotide kinase to give 12a. The 19-mer was also 3′-end labeled either with [R-32P]ddATP and terminal transferase or by end filling with [R-32P]dATP and Klenow fragment in the presence of a complementary strand with a 5′-protruding T to give 12b. Exposure of the labeled oligodeoxynucleotides to increasing amounts of 366 nm irradiation from a hand-held UV-lamp caused increasing amounts of strand cleavage. With the 5′-end labeled 12a, only a single cleavage band was produced that comigrated with the expected 3′-phosphate terminated product that was produced by a Maxam-Gilbert G sequencing reaction (20) on a substrate with a G in place of Y (Lane G, Figure 3b). With the 3′-end labeled substrate 12b, three major cleavage products were observed (Figure 3c) as expected from the results of the previous study on the photolysis of 2 (Figure 2b) (17). When the product mixture resulting from 16 min of photolysis was further treated with hot piperidine, the slower moving bands disappeared and only the band comigrating with the 5′-phosphate-terminated product was observed (Lane P, Figure 3c), also as expected from the results of the prior study. Now that we had a caged strand break Y that could produce a 3′-phosphate directly upon photolysis, we expected that we could directly produce a gap terminating in both 3′- and 5′-phosphates by incorporating caged strand breaks Y and X in tandem (Figure 4). To investigate this idea, we incorporated the caged strand breaks Y and X side by side into oligodeoxynucleotide 14 (n ) 0). Irradiation of either the 5′- or 3′-end labeled substrates 14a or 14b (n ) 0) with 366 nm light resulted in strand cleavage, ultimately giving only one band which comigrated with an authentic reference band produced by a Maxam-Gilbert G sequencing reaction (Lane G, Figure 5c). This band was also resistant to further treatment with hot piperidine (Lane P, Figure 5c). Unexpectedly, very little of the expected intermediate band corresponding to 15b or 16 was observed by visual inspection of the autoradiograms, suggesting that the second photocleavage event might be faster than the first photocleavage event. It was possible that an intermediate o-nitroso ketone product from the first photocleavage event, which is known to absorb strongly at 366 nm, or some other product might be sensitizing the second photocleavage event. If true, such a photosensitization mechanism should be distance dependent, and oligodeoxynucleotides were also synthesized in which the caged strand breaks Y and X were spaced 1 and 4 nucleotides apart (14, n ) 1 or 4). These substrates were also of interest because they could be used to prepare caged DNA strand gaps when annealed to a complementary strand. Plotting the relative yield of reaction products quantified by Phosphoimager analysis as a function of irradiation time



Figure 5. Expected products of the photolysis of (a) 5′- and (b) 3′-end-labeled oligodeoxynucleotides 14a and 14b containing two tandem caged single strand breaks X and Y. To the right are the autoradiograms of denaturing polyacrylamide electrophoresis gels of the products resulting from exposure of the corresponding (c) 5′- and (d) 3′-end labeled oligodeoxynucleotides 14a and 14b (n ) 0) to the indicated number of minutes of 366 nm light from a hand held UV-lamp. The products of the 16 min irradiation were also subjected to 1 M piperidine at 90 °C for 30 min (lane P). The fastest moving band in panel c corresponds to the product 5 that was produced by a Maxam-Gilbert G reaction that was carried out on an otherwise identical sequence in which X and Y was replaced with a single pdG (lane G). Lane S corresponds to an authentic sample of 5′-end labeled 4b.

(Figure 6) revealed that the rate of formation and decay of the intermediate band corresponding to 16 was the same for each spacing, indicating that the photosensitization mechanism was not taking place. The photolysis reactions were also repeated in the presence of 5 mM DTT to react with and eliminate a possible nitroso group (21), but once again, the rate of formation and decay of the intermediate band corresponding to 16 was the same as in the previous experiment (data not shown). The maximum value of approximately 25% of the intermediate is consistent with a mechanism in which the rates of cleavage at either site are identical and independent of the order of cleavage, as determined by simulation of the kinetics. To demonstrate the formation of a double-strand break in which one of the duplex fragments is terminated by 5′- and 3′-phosphates, we took advantage of an internally labeled hairpin substrate used by Stubbe and co-workers to probe double strand break formation induced by bleomycin (22). Two 54 mer hairpins 18 and 19 were constructed (Figures 7a and 8a), in which the caged strand breaks were one and three nucleotides apart, respectively, on opposite strands. The internally 32Plabeled hairpins were constructed by ligation of a preformed 34-mer hairpin structure that had the caged strand break X 13 nucleotides from the 5′-end to either one of two different 5′-end labeled 15-mers in which the caged strand break Y was either 6 or 9 nucleotides from the 3′-end. Single strand cleavage at X could be differentiated from single strand cleavage at Y and from a

double strand break by the length of the radiolabeled oligodeoxynucleotides produced. Irradiation under native (Figure 7b and 8b) and denaturing conditions (Figure 7c and 8c) indicated that photocleavage was sensitive to secondary structure for the more closely spaced strand break substrate but not for the more separated one. For the more closely spaced strand break substrate 18, cleavage at Y was slower in the duplex than in the denatured state. In the more closely opposed arrangement, building block Y is stacked between a T and an A, whereas in the more separated arrangement it is stacked between a C and an A. In the denatured state, cleavage at Y was faster than cleavage at X for both substrates, which may reflect an intrinsic preference for photocleavage at a 3′-phosphate over a 5′-phosphate. In any case, almost complete double strand cleavage was observed without the formation of any side products within 16 min of irradiation, making these caged substrates synthetically useful. CONCLUSION

We have developed a new building block for caging sitespecific and sequence-independent DNA strand breaks leading to the direct production of a 3′-phosphate, with doses of UVA light that cause little or no damage to DNA or protein, and that are not significantly lethal to mammalian cells (23). We have also demonstrated that when this building block is linked in tandem with our previously designed building block (17), it can be used to generate a major type of DNA strand break induced



Figure 6. Plots of the photolysis products of 5′-end labeled oligonucleotide 14a (n ) 0, 1, and 4) as a function of irradiation time.

by ionizing radiation, a nucleotide gap terminating in 3′and 5′-phosphates. These caged single strand breaks should be useful in the construction of DNA fragments and phagemid-based substrates needed to study the structure-activity relationships of single and double strand break repair using reconstituted repair systems or cell free extracts. The caged breaks could be introduced into phagemids by standard methods used to incorporate DNA damage-containing oligodeoxynucleotides into single and double-stranded vectors (24, 25) and then decaged following purification. To be a useful “caged” substrate for in vivo studies of strand break repair and mutagenesis, the modified oligonucleotides must not be substrates for other repair processes, such as nucleotide excision repair (NER), that might repair them before they can be activated. Although NER is generally attributed to repair of bulky lesions, it has been shown to repair DNA containing synthetic mimics of abasic sites (26). If the two caged breaks Y and

X directly follow each other, they may create a highly distorted DNA structure that might be readily detected and repaired by NER. The two building blocks might not be as easily detected and repaired, however, if they were spaced apart. Thus, to form a double strand break terminating in 3′- and 5′-phosphates in vivo, it may be better to space the two building blocks apart in each strand. Irradiation of such a construct would then produce a short duplex fragment in addition to the desired phosphate terminated double strand break. For in vivo studies, it may also be possible to use a single caged break X or Y in each strand, because endogenous exonucleases may be able to remove the X′ or Y′ that remains on the 3′- or 5′-phosphate following photolysis (Figures 2a and 3a). For instance, in E. coli, exonuclease III is known to process 3′-phosphate, 3′-phosphoglycolate, and other fragmented sugars groups left on the 3′-end (8). Clearly, further studies are required to explore the usefulness of these caged substrates.



Figure 7. (a) Expected single and double strand cleavage products resulting from the photolysis of the internally labeled hairpin substrate 18 containing caged strand breaks X and Y that are one nucleotide apart. Below are plots of the product distributions as a function of irradiation time under (b) native and (c) denaturing conditions.

Figure 8. (a) Expected single and double strand cleavage products resulting from the photolysis of the internally labeled hairpin substrate 19 containing caged strand breaks X and Y that are three nucleotides apart. Below are plots of the product distributions as a function of irradiation time under (b) native conditions and (c) denaturing conditions. ACKNOWLEDGMENT

This work was partially supported by NIH Grants CA51116 and CA40463 and a GANN fellowship to P.O. (Department of Education Grant P200A20106). The assistance of the Washington University High Resolution NMR Facility, funded in part through NIH Biomedical Research Support Shared Instrument Grants RR-02004,

RR-05018, and RR-07155 is gratefully acknowledged, as is the Washington University Mass Spectrometry Resource, an NIH Research Resource (Grant P41RR0954). LITERATURE CITED (1) Sonntag, C. v. (1987) The Chemical Basis of Radiation Biology, Taylor & Francis, New York.

Caged Single and Double Strand Breaks (2) Obe, G., Johannes, C., and Schulte-Frohlinde, D. (1992) DNA double-strand breaks induced by sparsely ionizing radiation and endonucleases as critical lesions for cell death, chromosomal aberrations, mutations and oncogenic transformation. Mutagenesis 7, 3-12. (3) Ward, J. F. (1995) Radiation mutagenesis: the initial DNA lesions responsible [published erratum appears in Radiat. Res. (1995) 143 (3), 355]. Radiat. Res. 142, 362-368. (4) Povirk, L. F. (1996) DNA damage and mutagenesis by radiomimetic DNA-cleaving agents: bleomycin, neocarzinostatin and other enediynes. Mutat. Res. 355, 71-89. (5) Dedon, P. C., and Goldberg, I. H. (1992) Free-radical mechanisms involved in the formation of sequence-dependent bistranded DNA lesions by the antitumor antibiotics bleomycin, neocarzinostatin, and calicheamicin. Chem. Res. Toxicol. 5, 311-332. (6) Thacker, J. (1994) The study of responses to “model” DNA breaks induced by restriction endonucleases in cells and cellfree systems: achievements and difficulties. Int. J. Radiat. Biol. 66, 591-596. (7) Giese, B., Dussy, A., Elie, C., Erdmann, P., and Schwitter, U. (1994) Synthesis and selective radical cleavage of C-4′modified oligonucleotides. Angew. Chem., Int. Ed. Engl. 33, 1861-1863. (8) Friedberg, E. C., Walker, G. C., and Siede, W. (1995) DNA Repair and Mutagenesis, ASM Press, Washington DC. (9) Marriott, G. (1998) Caged compounds. Methods in Enzymology, Academic Press, New York. (10) Engels, J., and Schlaeger, E. J. (1977) Synthesis, structure, and reactivity of adenosine cyclic 3′,5′-phosphate benzyl triesters. J. Med. Chem. 20, 907-911. (11) Hoffman, J. F., Forbush, B., III, and Kaplan, J. H. (1978) Rapid photolytic release of adenosine 5′-triphosphate from a protected analogue: utilization by the Na:K pump of human red blood ghosts. Biochemistry 17, 1929-1935. (12) Horn, T., Downing, K., Gee, Y., and Urdea, M. S. (1991) Controlled chemical cleavage of synthetic DNA at specific sites. Nucleosides Nucleotides 10, 299-302. (13) Kotera, M., Bourdat, A.-G., Defrancq, E., and Lhomme, J. (1998) A highly efficient synthesis of oligodeoxyribonucleotides containing the 2′-deoxyribonolactone lesion. J. Am. Chem. Soc. 120, 11810-11811. (14) Chaulk, S. G., and MacMillan, A. M. (1998) Caged RNA: photocontrol of a ribozyme reaction. Nucleic Acids Res. 26, 3173-3178. (15) Nadji, S., Wang, C.-I., and Taylor, J.-S. (1992) Photochemically and photoenzymatically cleavable DNA. J. Am. Chem. Soc. 114, 9266-9269.

Bioconjugate Chem., Vol. 11, No. 1, 2000 103 (16) Giese, B., Beyrich-Graf, X., Erdmann, P., Giraud, L., Imwinkelried, P., Muller, S. N., and Schwitter, U. (1995) Cleavage of single-stranded 4′-oligonucleotide radicals in the presence of oxygen. J. Am. Chem. Soc. 117, 6146-6147. (17) Ordoukhanian, P., and Taylor, J.-S. (1995) Design and synthesis of a versatile photocleavable DNA building block. Application to phototriggered hybridization. J. Am. Chem. Soc. 117, 9570-9571. (18) Fasman, G. D. (1975) Nucleic Acids. Handbook of Biochemistry and Molecular Biology, Vol. I CRC Press, Cleveland. (19) Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd ed., Vol. 1, Cold Spring Harbor Laboratory Press, Plainview, NY. (20) Maxam, A. M., and Gilbert, W. (1980) Sequencing endlabeled DNA with base-specific chemical cleavages. Methods Enzymol. 65, 499-560. (21) Trentham, D. R., Corrie, J. E. T., Barth, A., Gradwell, M. J., Maeda, Y., Mantele, W., and Meir, T. (1997) Time-resolved infrared spectroscopy of intermediates and products from photolysis of 1-(2-nitrophenyl)ethyl phosphates: reaction of the 2-nitrosoacetophenone byproduct with thiols. J. Am. Chem. Soc. 119, 4149-4159. (22) Stubbe, J., Absalon, M. J., and Kozarich, J. W. (1995) Sequence-specific double-strand cleavage of DNA by Febleomycin. The detection of sequence-specific double-strand breaks using hairpin oligonucleotides. Biochemistry 34, 20652075. (23) Churchill, M. E., Peak, J. G., and Peak, M. (1991) Correlation between cell survival and DNA single-strand break repair proficiency in the chinese hamster ovary cell line AA8 and EM9 irraidated with 365-nm ultraviolet-A radiation. Photochem. Photobiol. 53, 229-236. (24) Basu, A. K., and Essigmann, J. M. (1988) Site-specifically modified oligodeoxynucleotides as probes for the structural and biological effects of DNA-damaging agents. Chem. Res. Toxicol. 1, 1-18. (25) Banerjee, S. K., Christensen, R. B., Lawrence, C. W., and LeClerc, J. E. (1988) Frequency and spectrum of mutations produced by a single cis-syn thymine-thymine cyclobutane dimer in a single-stranded vector. Proc. Natl. Acad. Sci. U.S.A. 85, 8141-8145. (26) Huang, J. C., Hsu, D. S., Kazantsev, A., and Sancar, A. (1994) Substrate spectrum of human excinuclease: repair of abasic sites, methylated bases, mismatches, and bulky adducts. Proc. Natl. Acad. Sci. U.S.A. 91, 12213-12217.

BC9900993

Caged Single and Double Strand Breaks - Bioconjugate Chemistry

Recommend Documents