Proc. Nati. Acad. Sci. USA Vol. 88, pp. 8835-8839, October 1991 Chemistry Studies on DNA-cleaving agents: Computer modeling analysis of the mechanism of activation and cleavage of dynemicin- oligonucleotide complexes PAUL A. WENDER*, ROBERT C. KELLYt, SUZANNE BECKHAM, AND BENJAMIN L. MILLER Department of Chemistry, Stanford University, Stanford, CA 94305 Communicated by John I. Brauman, July 15, 1991 (receivedfor review May 13, 1991)

ABSTRACT Dynemicin A is a recently identified antitu- 506, version 2.1) with the AM1 Hamiltonian (7).] Abstraction mor antibiotic. Upon activation, dynemicin is reported to cause ofproximate deoxyribosyl hydrogens by this diradical would double-stranded cleavage of DNA, putatively through the initiate oxidative cleavage on opposing DNA strands. Con- intermediacy of a diradical. Computer modeling of this acti- version of diradical 4 to the alternative ene-diyne structure 7 vation and cleavage process is described herein as part of an is not observed. effort to establish a structural hypothesis for this mechanistic Semiempirical (8) and molecular mechanics (9) studies on sequence and for the design of simple analogues. Intercalation dynemicin itself have provided valuable information in sup- complexes of duplex dodecamers [d(CGCGAATTCGCG)J2 port of the above mechanism. Thus far, however, computa- and [d(GC)6]2 with both enantiomers of dynemicin and of all tional methods have not been used to evaluate the role of related mechanistic intermediates are evaluated. Examination DNA in the mode of action ofdynemicin, although they have of these structures shows that cycloaromatization of dynemicin been applied to (10) and (11), to a diradical intermediate results in the rotation of the yielding models that are consistent with known DNA cleav- diradical-forming subunit with respect to the intercalation age patterns. We describe herein computer modeling studies plane that is of an opposite sense for the two dynemicin designed to delineate at the molecular level the interaction of enantiomers. In addition, the activation of the (2S) enantiomer dynemicin and dynemicin-derived intermediates with oligo- of dynemicin occurs by a less restricted approach trajectory nucleotides selected to emulate native DNA. These studies than (2R) enantiomer. In all complexes, the address several fundamental issues that are crucial to the the corresponding development ofa structural hypothesis for the mode ofaction 5'-3' strand is at least 1 A closer than the 3'-5' strand to the of dynemicin and its analogues, including (i) the mechanistic diyl intermediate. As a result, complexes are produced in which fate ofthe two possible enantiomers of dynemicin [structures the diyl moiety is aligned along [(2S)J or across [(2R)] the minor 1-(2R) or 1-(2S)], a point of much interest since the absolute groove, leading to different predictions for the selectivity of stereochemistry of dynemicin is as yet unknown, (ii) the radical-initiated, oxidative lesion of DNA. Molecular dynamics effect of nucleophile size and approach trajectory in the simulations are found to support these predictions, including activation step and the dynamics of this activation process, the 3-base-pair offset cleavage reported for dynemicin. (iii) the influence of oligonucleotide sequence and length on dynemicin intercalation and activation, and (iv) the relation- The cleavage of DNA is a key process in the transfer of ship of intercalation sites to cleavage sites. The answers to genetic information, the mode of action of certain chemo- these questions provide a structural basis for evaluating therapeutic agents, and the function of reagents designed for mechanistic proposals, for predicting DNA cleavage pat- DNA modification and structure determination. DNA cleav- terns, and for designing new cleaving agents based on the age can be effected with a variety of agents ranging from the dynemicin lead. simple hydroxyl radical to relatively complex restriction enzymes. Within the past 5 years, the antitumor antibiotics calicheamicin (1), (2), and neocarzinostatin (3) METHODS have emerged as a new structural and mechanistic class of A DNA octamer corresponding to [d(CGAATTCG)L2 and DNA-cleaving agents that are proposed to operate through dodecamers corresponding to [d(CGCGAATTCGCG)]2 and the inducible generation ofan arenyl or indenyl diradical. The [d(GC)62 were constructed in B-DNA form by using the most recently identified member of this class is dynemicin A program MACROMODEL (versions 2.0 and 3.0; W. C. Still, (structure 1 in Scheme I), a compound that exhibits potent Columbia University) running on a MicroVax 3900 and Evans cytotoxicity and in vivo antitumor activity (4, 5). Dynemicin and Sutherland PS340 system. All sequences were minimized has been shown to interact with the minor groove of DNA to a gradient of <0.100 kJ/mol per A under the AMBER force and, upon activation, to cause double-strand breaks 3 base field (12) before intercalation experiments were begun. Both pairs (bp) apart (6). Examination of the structure of dynemi- enantiomers corresponding to dynemicin A 1, the putative cin suggests that it could be activated for DNA cleavage quinone methide intermediate 2, the proposed quinone me- through reduction of its subunit, resulting in thide addition product 3, and a surrogate for the cyclized, heterolysis of the adjacent epoxide ring. Addition of a nu- diradical intermediate 5 were minimized to a gradient of cleophile to, or protonation of, the resultant anthraquinone <0.200 kJ/mol per A by using the MM2 force field (13, 14). methide (structure 2 in Scheme I) would provide an activated A surrogate for diradical intermediate 4 was necessary be- derivative 3 which, in the absence ofthe constraints imposed cause current molecular mechanics force fields are not pa- by the original epoxide ring, would undergo facile cycloar- rameterized for diradicals; the pyrazine ring was chosen as a omatization to diradical 4 (Scheme I). [Heats of formation surrogate because of its geometric similarity to the putative were determined by using the AMPAC program (QCPE no. diradical species. As the current parameter set available in

The publication costs of this article were defrayed in part by page charge *To whom reprint requests should be addressed. payment. This article must therefore be hereby marked "advertisement" tSenior Scientist on leave from the Upjohn Company, Kalamazoo, in accordance with 18 U.S.C. §1734 solely to indicate this fact. MI 49001.

8835 Downloaded by guest on October 2, 2021 8836 Chemistry: Wender et al. Proc. Natl. Acad. Sci. USA 88 (1991)

25 24 0 OH 0 HN O,-89.261 wlroH OH 0 HN OH '- - OCH3 - -89.261 kcaVmol -125.574 kcaVrmol OH 0 OH OH OH OH 2 NOHB C 3 N N 0 0 0 OH OH OH HN OH OH OH HNHOP OH OH OH HNHO OH - NN E CH3 -, N N OCH3 p H H > H -147.107 kcaVmol -131.156 kcaVmol OH OH OH OHOH OH OH OH OH 3 4 5

0 0 OH OH HNHO OH OH OHHNHO OH NCOCHH 3 H I H -1 55.887 kcaVmol -255.815 kcal/mol OH OH OH OH OH OH 7 6 Scheme I

AMBER did not extend to acetylenes, it was necessary to H-N-N angle. The H-lone pair-N angle was obtained from introduce standard parameters from MM2 into AMBER. In the this value, and the measured N-H distance by triangulation, case of stretching interactions, this was done by multiplying where the N-lone pair distance was set to 0.600 A, the value the MM2 constant by a proportionality factor; with other given in the MM2 force field (13, 14). parameters it was possible to use them without modification, Molecular dynamics simulations were carried out by using or to use values for available substructures in AMBER which the minimized complexes of 5-(2R) and 5-(2S) with [d(CGC- had force constants in MM2 identical to those of acetylenes. GAATTCGCG)h2 as starting structures. Bonds to hydrogen The resulting modified AMBER force field gave structures that were constrained with the SHAKE algorithm (16), and the were identical to those obtained with MM2 and, for dyne- thermal stability of each system was maintained by coupling micin A itself, produced a structure consistent with the x-ray to a 300 K external bath (17). Following a 15-ps preequili- crystallographic structure (5). (Copies of the modified pa- bration, the complexes were observed for 30 ps. A timestep rameters used in these calculations are available from the of 1 fs was used during both the preequilibration and obser- authors upon request.) vation periods, and the nonbonded interaction array was Initial intercalation spaces in the duplex oligonucleotides updated every 0.5 ps. were formed by docking only the anthraquinone portion of dynemicin into position between base pairs and minimizing for 1000 iterations under AMBER. The intercalator was then RESULTS removed, and the molecule of interest was placed into the Because of the absence of DNA cleavage data at the outset gap. The entire structure was then minimized to a gradient of ofthis study, G+C- and A+T-rich oligonucleotide sequences <0.200 kJ/mol per A by using the modified force field. were selected to explore two generic intercalation sites. Structures used to study the effect of nucleophile size on Sequence selection around these sites was guided by sym- DNA structure and strain energy were generated by adding metry considerations (to simplify calculations) and/or by the a hydrogen at C8 of 2 in the minimized complex, rearranging availability of solid state (18) or solution phase (19-24) the bond orders to give an unminimized form of 3, and structural information (for calibration). Sequence length was carrying out a substructure minimization on C8 to permit that initially set at 8 bp to minimize computational time. How- atom to pyramidalize. This yielded a complex designed to ever, an early finding of this study was that the intercalation mimic the conditions that would exist immediately after complexes ofthe resultant duplex octamers exhibit disrupted addition of a nucleophile. The hydrogen at C8 was then base pairing at terminal residues when full, rather than replaced by a methyl, ethyl, isopropyl, or tert-butyl group. substructure, minimizations were conducted. Since this mod- Each complex was then minimized as described above, and eling suggested, in agreement with solution studies of DNA the strain energy was calculated. intercalation complexes, that the effects of intercalation are Minor groove widths were determined by measuring the transmitted more than four residues in both directions from perpendicular distance between the centers of mass of phos- the site of intercalation, the oligonucleotide length was ex- phorus atoms on complementary strands and subtracting 5.8 tended to duplex dodecamers. The resultant duplex structure A, the Van der Waals diameter of a phosphate group (15). for uncomplexed [d(CGCGAATTCGCG)h2 was found to dif- Angles for hydrogen abstraction (N-lone pair-H angles) for fer only slightly from the x-ray crystal structure (18), the the pyrazine diyl surrogate complexes were calculated by latter being slightly more tightly wound as a result of crystal- assuming the pyrazine ring to be planar and measuring the packing forces. Intercalation complexes ofboth duplex dode- Downloaded by guest on October 2, 2021 Chemistry: Wender et al. Proc. Natl. Acad. Sci. USA 88 (1991) 8837 camers produced a DNA model that displayed normal end group base pairing. Since the absolute stereochemistry of dynemicin has not been established, intercalation complexes were constructed of the aforementioned duplex dodecamers with both possible enantiomers of dynemicin and of the resultant quinone meth- ide, cyclization precursor, and pyrazine surrogate for the diradical. All minimized duplexes were found to be in the right-handed, Watson-Crick B form, with most of the sugar ring puckers C2'-endo. The overall structure of the oligonu- cleotides in the complexes was little changed from the uncom- plexed form. Separation of the base pairs immediately above and below the intercalation site increased from 3.4 A to 6.4-6.6 A upon intercalation ofdynemicin, and Watson-Crick base pairing remained intact. The intercalation site was wedge- shaped, similar to that observed for a nogalamycin- oligonucleotide complex (25) and a daunomycin-oligonucleo- tide complex (26). Base pairs directly adjacent to the interca- lation site exhibited a slight propellor twist and were buckled. The anthraquinone moiety was puckered and not planar. Although this puckering is not observed in the daunomycin FIG. 2. (2S)-diyl surrogate (Left) and (2R)-diyl surrogate (Right) complex, it is seen in nogalamycin, which, like dynemicin, has in [d(CGCGAA1TCGCG)]2- an aliphatic ring rigidly attached to the intercalator subunit (25). corresponding (2S)-diyl surrogate complex (Figs. 2 Left and Two general energy minima were found for the complexes, 3 Left), atoms C23-C28 move in a clockwise sense with one with the anthraquinone intercalated more deeply and the respect to the pseudo-plane ofthe intercalated anthraquinone ene-diyne moiety closer to the double-helical axis than in the subunit, resulting in a structure in which these atoms are other. The proposed explanation for this is that H5' and/or aligned along the minor groove. The initial alignment of the H4' of base 8 cause unfavorable Van der Waals interactions (2R)-dynemicin complex (Fig. 1 Right) is similar to that found when they are in the same plane as the ene-diyne (or diradical for the (2S) complex. However, as expected from the con- surrogate) portion. This results in a local minimum when the trasting handedness of the (2R)-dynemicin isomer, its con- ene-diyne/diradical surrogate is inserted beyond these hy- version to the (2R)-diyl surrogate proceeds with the coun- drogens. Although the size of the barrier between these two terclockwise movement of atoms C23-C28 with respect to minima is unknown, experimental data available for the the intercalated anthraquinone subunit, producing a structure structurally related daunomycin-oligonucleotide complex, in in which these atoms align across the minor groove (Figs. 2 which the amino sugar lies on the floor of the minor groove Right and 3 Right). and the D ring protrudes into solution from the major groove, Although alignment of the intercalating species along or suggests that deeply intercalated species are energetically across the minor groove does not by itselfdetermine whether accessible. This also suggests that intercalation depth may be the resultant diyl could cleave DNA, it does determine access an important structural parameter in the design of analogues to and steric encumbrance, at least for nucleophilic attack, at for abstraction of specific hydrogens. the site of attachment for the incoming nucleophile or proton Comparison ofthese complexes provides an understanding source. This is clearly seen for the (2S) derivatives, where the of the structural changes that attend the activation and trajectory of nucleophilic attack or protonation would be cyclization steps, including the differing roles of the two easily accommodated along the minor groove, while the path possible dynemicin enantiomers. Specifically, for the con- for the (2R) derivatives would encounter severe steric en- version of the (2S)-dynemicin complex (Fig. 1 Left) to the

FIG. 3. Corey-Pauling-Koltun models of the (2S)-diyl surrogate FIG. 1. (2S)-dynemicin (Left) and (2R)-dynemicin (Right) in (Left) and (2R)-diyl surrogate (Right) in [d(CGCGAATTCGCG)]2. [d(CGCGAATrCGCG)]2. The closest hydrogen for each static complex is highlighted in red. Downloaded by guest on October 2, 2021 8838 Chemistry: Wender et al. Proc. Natl. Acad. Sci. USA 88 (1991) 30-. (2S) bases 1-12 30 25- 25 20 20 666 6 15 15 10 000. 10 BasseS.HI |l.a I 9Angstm a

E OH - B.1 B2 B3 B4 B5 B6 B7 B8 B9 B10 Bll B12 5 31 I 5'0 .1 5'% sU (4)MONaLoases--- s 13-24.2e % A Cud ._ 2

zw A1ill;.itI .I.5 6!

l ' 7 5 .Hii i i Angstrolillu§§ins .-.ul28@@|§ell. 0 B13 B14 B15 B16 B17 B18 B19 B20 B21 B22 B23 B24 B13 B14 B15 B16 B17 B18 B19 B20 B21 B22 B23 B24 5' O3' St - 3' FIG. 4. Maximal (x), minimal (o), and average (-) N-H5', N-H4', N-H3', and N-Hi' distances for 5 in [d(CGCGAATTCGCG)h2 from molecular dynamics simulations. cumbrance from the 3'-5' strand of the oligonucleotide. This provides an indication of the manner in which cleavage sites trend is further supported by the relative stabilities of the are related to intercalation sites and suggests where the initial activated complexes: in all cases, the (2S) complex is more abstraction might occur. It was found that the 5'-3' strand stable than the (2R) complex, with energy differences ranging was, on average, greater than 1 A closer to the "diyl" than from 8.82 kcal/mol for nucleophiles similar to a methyl group the 3'-5' strand in all of the static cases examined. For both in size to 14.72 kcal/mol for a tert-butyl group. For an (2S)-diyl complexes, the hydrogens on base 8 are closest. activation process involving protonation ofan anthraquinone (The optimum transition-state geometry for abstraction of methide, the product complexes are of comparable energy, hydrogen from methane by methyl radical has been deter- suggesting that either could be accommodated in the minor mined by ab initio methods; see ref. 27.) groove. However, protonation of the (2R)-anthraquinone Molecular dynamics simulations ofthe complexes of5-(2R) subunit would preferentially occur prior to association with and 5-(2S) with [d(CGCGAATTCGCG)J2 at 300 K confirm DNA, since direct protonation of the anthraquinone subunit the above general observations on the static complexes and when complexed with DNA would be sterically disfavored. provide additional information with respect to the differing Alternatively, a relay proton transfer mechanism could be interaction of the two enantiomers with DNA. While base 7 involved, in which case a small proton transfer agent (e.g., is once again closest to the "diyl" moiety of 5-(2R) through- water) could associate with DNA prior to the intercalation of out the simulation, the contrasting handedness of 5-(2S) dynemicin and might remain present in the minor groove so coupled with motion within the intercalation site causes base as to allow direct contact with dynemicin. 8 to be closest in the 5-(2S)-DNA complex (Fig. 4). None of An analysis of distances between abstractable hydrogens the bases ofthe 3'-5' strand come within abstraction distance on the sugar-phosphate backbone and radical centers on 4 of the 5-(2R)-"diyl" during the simulation, the closest being .3 y _ [d(GC)6J2 -2S-1+DNA ia(C)P- -C(G) 13 -- 2R-1+DNA Pairc#,4mPair#121 -2S-+DNA etc. i'!!P ,'~~~~~a)~~~~~~ 11 to Pair#1 E ,..~~..2R-5+DNA (C! 2

U s\, ~~~~6(cQy---(G) If ._O M4 (ERG- -C) _ ~ ~~~~~~~~~1-(Gc-_C X -y 2 (C)G'--(G)23

3 G 2 1 2 3 4 5 6 7 8 1 2 3 4 6 e0 7 Pair Number FIG. 5. Groove widths for [d(CGCGAATTCGCG)h2 and [d(GC)6]2 complexes. 2S-1, 2R-1, 2S-5, and 2R-5, (2S) and (2R) enantiomers of structures 1 and 5 of Scheme I. Downloaded by guest on October 2, 2021 Chemistry: Wender et al. Proc. Natl. Acad. Sci. USA 88 (1991) 8839 base 18 at an approximate minimum distance of 4 A. Bases tion from the 3'-5' strand would necessarily follow abstrac- 18 and 19 approach within 2 and 3.5 A, respectively, of the tion from the 5'-3' strand for the complex with 4-(2R). 5-(2S)-"diyl". Plots ofgroove widths measured for each dynemicin model Support of this research by the National Cancer Institute through intercalated into [d(CGCGAATTCGCG)12 and for the free Grant CA31845 is gratefully acknowledged. S.B. is a National dodecamer are shown in Fig. 5. As expected from the Science Foundation Research Fellow, and B.L.M. is a W. R. Grace "wrong-handed" twist observed on cyclization of 3-(2R), Graduate Fellow. of 5-(2R) its diyl model caused greater intercalation and 1. Zein, N., Sinha, A. M., McGahren, W. J. & Ellestad, G. A. broadening of the minor groove than intercalation of 5-(2S). (1988) Science 240, 1198-1201. The larger groove widths observed for the (2R) enantiomer 2. Long, B. H., Golik, J., Forenza, S., Ward, B., Rehfuss, R., relative to the (2S) enantiomer in [d(CGCGAATTCGCG)]2 Dabrowiak, J. C., Catino, J. J., Musial, S. T., Brookshire, systems are not observed for the GC dodecamer [(GC)612 K. W. & Doyle, T. W. (1989) Proc. Nati. Acad. Sci. USA 86, (Fig. 5). It has been observed (15) in x-ray diffraction studies 2-6. on DNA fibers that the minor groove width of B-DNA 3. Napier, M. A., Holmquist, B., Strydom, D. J. & Goldberg, increases with the proportion of G-C base pairs, and it is I. H. (1979) Biochem. Biophys. Res. Commun. 89, 635-642. 4. Konishi, M., Ohkuma, H., Matsumoto, K., Tsuno, T., Kamei, probable that the larger minor groove width of the GC H., Miyaki, T., Oki, T., Kawaguchi, H., VanDuyne, G. D. & dodecamer permits intercalation with less distortion. Clardy, J. (1989) J. Antibiot. 42, 1449-1452. 5. Konishi, M., Ohkuma, H., Tsuno, T., Oki, T., VanDuyne, G. D. & Clardy, J. (1990) J. Am. Chem. Soc. 112, 3715-3716. CONCLUSION 6. Sugiura, Y., Shiraki, T., Konishi, M. & Oki, T. (1990) Proc. Sixteen intercalation complexes ofboth possible enantiomers Nati. Acad. Sci. USA 87, 3831-3855. of dynemicin and of dynemicin-derived intermediates with 7. Dewar, M. J. S., Zoebisch, E. G., Healy, E. F. & Stewart, J. J. P. (1985) J. Am. Chem. Soc. 107, 3902-3909. duplex octamers and dodecamers were constructed and 8. Snyder, J. P. & Tipsword, G. E. (1990) J. Am. Chem. Soc. 112, minimized. While the duplex octamer complexes suffered 4040-4042. from disrupted base pairing at the termini, complexes with 9. Semmelhack, M. F., Gallagher, J. & Cohen, D. (1990) Tetra- duplex dodecamers exhibited normal end group base-pairing. hedron Lett. 31, 1521-1522. This suggests that duplex octamers may not be a reliable 10. Hawley, R. C., Kiessling, L. L. & Schreiber, S. L. (1989) model for DNA in dynemicin-oligonucleotide complexes, Proc. Nati. Acad. Sci. USA 86, 1105-1109. and longer oligonucleotides should be used. Both enantio- 11. Galat, A. & Goldberg, I. H. (1990) Nucleic Acids Res. 18, 2093-2099. mers of dynemicin and putative precyclization intermediates 12. Weiner, S. J., Kollman, P. A., Case, D., Singh, U. C., Ala- yield intercalation complexes with similar orientations of the gona, G., Profeta, S. & Weiner, P. (1984) J. Am. Chem. Soc. ene-diyne within the minor groove of DNA. Upon examina- 106, 765-784. tion ofthe trajectory ofattack ofa nucleophile to the quinone 13. Allinger, N. L. (1977) J. Am. Chem. Soc. 99, 8127-8134. methide intermediate, however, differences between the two 14. Burkert, U. & Allinger, N. L. (1982) Molecular Mechanics enantiomers become apparent. The approach of a nucleo- (Am. Chem. Soc., Washington). phile to either enantiomer is hindered by the 3'-5' strand of 15. Arnott, S. (1981) in Topics in Nucleic Acid Structure, ed. and the effect is greater for the (2R) enantiomer. In Neidle, S. (Halsted, New York), pp. 65-82. DNA, 16. Ryckaert, J.-P., Ciccotti, G. & Berendsen, H. J. C. (1977) J. addition to this kinetic effect on activation, it is also found Comput. Phys. 23, 327. that the activated (25) complex is lower in energy than the 17. Berendsen, H. J. C., Postma, J. P. M., van Gunsteren, W. F., (2R) complex for various nucleophiles. Cyclization to the diyl DiNola, A. & Haak, J. R. (1984) J. Chem. Phys. 81, 3684-3690. intermediate results in a rotation of the diyl moiety with 18. Drew, H. R., Wing, R. M., Takano, T., Broka, C., Tanaka, S., respect to the intercalation plane, improving the alignment of Itakura, K. & Dickerson, R. E. (1981) Proc. Natl. Acad. Sci. the (2S) enantiomer within the minor groove, while causing USA 78, 2179-2183. the (2R) enantiomer to become aligned across the minor 19. Patel, D. J. (1976) Biopolymers 15, 533-558. 20. Uesugi, S., Shida, T. & Ikehara, M. (1981) Chem. Pharm. Bull. groove. Although it is not yet known whether dynemicin 29, 3573-3585. intercalates before or after activation occurs, the above 21. Cheng, D. M., Kan, L. S., Frechet, D., Ts'O, P. 0. P., Ue- evidence suggests that if intercalation occurs first, the (2S) sugi, S., Shida, T. & Ikehara, M. (1984) Biopolymers 23, enantiomer of dynemicin is better suited for nucleophilic 775-795. activation within the complex and also indicates that any 22. Patel, D. J., Kozlowski, S. A., Markey, L. A., Broka, C., nucleophile attacking at C8 of the quinone methide interme- Rice, J. A., Itakura, K. & Breslauer, K. J. (1982) Biochemistry diate must be small. 21, 428-436. In all complexes, the diyl is much closer to the 5'-3' strand 23. Patel, D. J., Pardi, A. & Itakura, K. (1982) Science 216, of the oligonucleotide than to the 3'-5' strand. Even so, 581-590. 24. Pardi, A., Morden, K. M., Patel, D. J. & Tinoco, I., Jr. (1982) significant motion ofthe intermediate within the intercalation Biochemistry 21, 6567-6574. complex would be expected to occur, permitting abstraction 25. Searle, M. S., Hall, J. G., Denny, W. A. & Wakelin, L. P. G. from the 3'-5' strand. Molecular dynamics simulations de- (1988) Biochemistry 27, 4340-4349. signed to address this issue suggest that, although motion 26. Wang, A. H.-J., Ughetto, G., Quigley, G. J. & Rich, A. (1987) within the intercalation site brings the 5-(2S)-diyl surrogate Biochemistry 26, 1152-1163. within abstraction distance of both strands of DNA, abstrac- 27. Wildman, T. A. (1986) Chem. Phys. Lett. 126, 325-329. Downloaded by guest on October 2, 2021