Monitoring the DNA Binding Kinetics of a Binuclear ... - ACS Publications

Fredrik Westerlund, L. Marcus Wilhelmsson, Bengt Norde´n, and Per Lincoln* ... T)2], in which the minor groove binding dye DAPI is used as an energy ...
0 downloads 0 Views 187KB Size
21140

J. Phys. Chem. B 2005, 109, 21140-21144

Monitoring the DNA Binding Kinetics of a Binuclear Ruthenium Complex by Energy Transfer: Evidence for Slow Shuffling Fredrik Westerlund, L. Marcus Wilhelmsson, Bengt Norde´ n, and Per Lincoln* Department of Chemistry and Bioscience, Chalmers UniVersity of Technology, SE-41296 Gothenburg, Sweden ReceiVed: June 27, 2005; In Final Form: August 31, 2005

The semirigid binuclear ruthenium complex ∆,∆-[µ-(11,11′-bidppz)(phen)4Ru2]4+ has been shown to rearrange slowly from an initial groove-bound nonluminescent state to a final intercalated emissive state by threading one of its bulky Ru(phen)2 moieties through the DNA base stack. When this complex binds to poly[d(AT)2], a further increase in emission from the complex is observed after completion of the intercalation, assigned to reorganization of the intercalated complex. We here report a study of the threading process in poly[d(AT)2], in which the minor groove binding dye DAPI is used as an energy transfer probe molecule to assess the distribution of ruthenium complex during and also after the actual threading phase. The emission from DAPI is found to change with the same rate as the emission from the ruthenium complex, and furthermore, DAPI does not disturb the binding kinetics of the latter, justifying it as a good probe of both the threading and the reorganization processes. We conclude from the change in the emission from both DAPI and the ruthenium complex with time that DAPI-ruthenium interactions are most pronounced during the process of threading of the complex, suggesting that the complexes are initially threaded slightly anticooperatively and thereafter redistribute along the DNA to reach their thermodynamically most favorable distribution. The final distribution is characterized by a small but significant binding cooperativity, probably as a result of hydrophobic interactions between the complex ions despite their tetravalent positive charges. The mechanism of “shuffling” the complex along the DNA chain is discussed, i.e., whether the ruthenium complex remains threaded (requiring sequential base-pair openings) or if unthreading followed by lateral diffusion within the ionic atmosphere of the DNA and rethreading occurs.

Introduction Small DNA-binding molecules have been of great interest to develop and study for the last 50 years because of their possible applications as diagnostic agents, genetic probes, or chemotherapeutics. Strong binding to DNA and slow dissociation from the binding sites have been proposed as two important properties of a good anticancer agent.1 Hence, molecules with a high DNA binding affinity and also a high activation barrier to dissociate from DNA are desirable. A naturally occurring molecule with such properties is the antitumor drug nogalamycin. It threads one of its two bulky sugar moieties through the DNA base stack, placing one moiety in each groove and the middle planar aromatic part of the drug intercalated, i.e., “sandwiched”, between the base pairs.2-8 Threading normally implies very slow DNA binding and dissociation kinetics. It has been suggested that the threading process involves local melting,9 elongation,10 and unstacking10 of the DNA-helix, as well as opening of at least one base pair. Furthermore, since the kinetically favored threading intercalation sites may not be the same as the ones with the highest thermodynamical stability, a subsequent reorganization, shuffling, may follow after binding to the initial intercalated state to the thermodynamically favored sites.2-4 Covalently linked dimers of the chiral complex [Ru(phen)2dppz]2+ have been found to resemble naturally occurring DNA binding molecules. One example is [µ-C4(cpdppz)2* Corresponding author. Telephone: +46-31-7723055. Fax: +46-317723858. E-mail: [email protected].

(phen)4Ru2]4+, which, like echinomycin,11-14 binds to DNA by bis-intercalation. Spectroscopic and kinetic analyses of this ruthenium dimer have shown that the molecule intercalates its extended heterocyclic aromatic dppz moieties, placing the two ruthenium centra in the same groove and the aliphatic linker in the opposite groove.15,16 Analysis of the binding and dissociation kinetics of the three stereoisomers of this complex lead to the conclusion that it is actually the two ruthenium moieties of the molecule, and not the aliphatic linker, that are threaded through the DNA base stack, requiring large and concerted transient conformational changes of DNA to reach the transition state.15,16 The parent compound [Ru(phen)2dppz]2+ has the same chromophore as the dimer described above, and its DNA binding and photophysical properties have been studied extensively.17-26 When bound to DNA, both [Ru(phen)2dppz]2+ and the dimer show an increase in fluorescence quantum yield by more than 3 orders of magnitude, a phenomenon referred to as the “lightswitch” effect.21,27 Hydrogen bonding between surrounding water molecules and the aza-nitrogens of the dppz ligand is believed to be involved in an efficient radiationless relaxation process of the excited state in aqueous solution. By protecting the aza-nitrogens from water, e.g. by intercalation of the dppz ligand between the bases of DNA, this relaxation path can be efficiently shut off and thus the emission quantum yield greatly increased. Another very interesting example of a threading intercalating dimer built up by two [Ru(phen)2dppz]2+units is the semirigid complex ∆∆-[µ-(11,11′-bidppz)(phen)4Ru2]4+ (∆∆-P4, Figure 1). The first publication on DNA binding of this molecule

10.1021/jp0534838 CCC: $30.25 © 2005 American Chemical Society Published on Web 10/19/2005

Redistribution of Threaded Ruthenium Complex

Figure 1. Structures of [µ-(11,11′-bidppz)(phen)4Ru2]4+ (a) and the probe molecule diamidinophenylindole (DAPI) (b).

reported it not to intercalate but instead to bind in one of the grooves of DNA, as evidenced from linear dichroism (LD) measurements on flow-oriented calf-thymus DNA (CT-DNA).28 However, later studies showed that ∆∆-P4 indeed intercalates eventually, by threading one of its bulky Ru(phen)2 moieties through DNA, but that the threading is extremely slow, needing weeks at room temperature with 10 mM NaCl or hours at 50 °C with 100 mM NaCl to be completed.29 The threading intercalation is accompanied by a large increase in fluorescence quantum yield, just as for the monomer.29 Both the threading process and the final binding mode for ∆∆-P4 is very similar to the ones for the natural antibiotic nogalamycin. The final binding mode of these dumbbell-like molecules has one bulky moiety in each groove and an aromatic part sandwiched between the bases of DNA. Comparing the rates of dissociation for ∆∆P4 and nogalamycin,3 it can be seen that they both dissociate very slowly, probably as a consequence of the requirement of similar transient structural changes of DNA. A crucial evidence for threading intercalation of ∆∆-P4 is that SDS-induced dissociation of the final binding mode in CT-DNA needs several hours at elevated temperature (50 °C) to dissociate, whereas a complex in the initial binding mode under the same conditions dissociates from DNA immediately.29,30 A close look at the emission increase characteristics for the threading intercalation of ∆∆-P4 to poly[d(A-T)2] has revealed that three exponentials are needed to fit the experimental data satisfactorily, one fast with a large increase in quantum yield and two progressively slower with lower amplitudes. This finding was surprising, since threading intercalation by basepair opening seems most likely to be the rate-determining step and subsequent rearrangements in the intercalation pocket should be much faster processes. However, evidence has recently been found that this kind of molecules can bind to DNA via a bimolecular process.31 Indeed, we show here that the binding of ∆∆-P4 to poly[d(A-T)2] can be fitted equally well with a bimolecular model, in which a single exponential is simply added to account for the very slow increase in emission. We here use fluorescence resonance energy transfer (FRET) from a probe donor molecule (DAPI, Figure 1) that is added during the threading process of ∆∆-P4 to poly[d(A-T)2] to study the threading and the subsequent slow rise in emission with time, when all ∆∆-P4 is supposed to be already intercalated. Exploiting the spectral overlap between DAPI emission and the ruthenium chromophore absorption, FRET has previously been used to study the binding of [Ru(phen)2dppz]2+ to DNA.32,33 In our study, we see changes in emission from DAPI on the same time scale as the emission from ∆∆-P4 changes, and we use the change to test the bimolecular model recently proposed31

J. Phys. Chem. B, Vol. 109, No. 44, 2005 21141

Figure 2. Association of ∆∆-P4 to poly[d(A-T)2] (gray) and fitting with three exponentials (black). Measurements done at 50 °C in aqueous buffer (4 mM HEPES, 2 mM MgCl2, 5 mM KCl at pH 7.4).

and also to probe the subsequent shuffling, a process earlier suggested for multi-intercalators34,35 and also in the case of DNA-nogalamycin interactions.4 Materials and Methods Chemicals. Poly[d(A-T)2] was purchased from Pharmacia Biotech and was used as obtained. DAPI was purchased from Sigma. ∆∆-P4 was synthesized as described elsewhere.30 All experiments were performed in aqueous buffers (4 mM HEPES, 2 mM MgCl2, 5 mM KCl at pH 7.4). Sample Preparation. Samples were prepared by mixing ruthenium complex and poly[d(A-T)2] dissolved in buffer. Unless otherwise stated, the poly[d(A-T)2] concentration was 80 µM. Concentrations were determined on a Varian Cary 4B spectrophotometer. The extinction coefficients used were 408nm ) 75 800 M-1 cm-1 for [µ-(11,11′-bidppz)(phen)4Ru2]4+, 262nm ) 6600 M-1cm-1 for poly[d(A-T)2], and 350nm ) 27 000 M-1 cm-1 for DAPI. Circular Dichroism. Circular dichroism (CD) is defined as the difference in absorbance of left and right circularly polarized light, CD ) Al - Ar. CD spectra were measured on a Jasco J-720 spectropolarimeter using a 1-cm quartz cell. All spectra were recorded between 220 and 650 nm and corrected for background contributions. Steady-State Luminescence. Emission spectra were recorded on a xenon lamp equipped SPEX Fluorolog τ-3 spectrofluorimeter (JY Horiba) between 370 and 800 nm using an excitation wavelength of 360 nm. Time-based association measurements were performed by setting the excitation wavelength at 360 nm, whereas the emission was monitored simultaneously at 450 and 630 nm using the S and T channel, respectively. The measurements were performed using a reduced cell with 4-mm path length, and the spectra were corrected for the inner filter effect by multiplying the observed intensity by 100.2DAPIc. Results Figure 2 shows the change in emission intensity that accompanies the threading intercalation of ∆∆-P4 to poly[d(AT)2] at 50 °C. Fitting the experimental curve satisfactorily requires an expression with three reciprocal rate constants of 100, 910, and 9130 s, respectively. The fastest apparent firstorder process produces a threaded intercalated species, proved by that addition of SDS at short times after mixing does not cause the fast dissociation expected for a nonthreaded complex.30 Thus, the two slower first-order processes have to be ascribed to subsequent rearrangements of already threaded complexes. However, we have recently shown that the enantiomer ΛΛ-P4 binds to poly[d(A-T)2] by a simple bimolecular rate law at

21142 J. Phys. Chem. B, Vol. 109, No. 44, 2005

Figure 3. Emission signal of DAPI integrated from 400 to 575 nm (top panel), and DNA-induced CD signal of DAPI, integrated from 325 to 410 nm (bottom panel) plotted as functions of total DAPI concentration (circles). ∆∆-P4 concentrations were 0, 2, 4, and 6 µM, and poly[d(A-T)2] concentration was 80 µM. The ∆∆-P4 concentration increases in the direction of the arrows. The lines are there only to guide the eye. Measurements were done at 50 °C in aqueous buffer (4 mM HEPES, 2 mM MgCl2, 5 mM KCl at pH 7.4). The emission spectra have been corrected for the inner filter effect.

high binding densities.31 Indeed, an equally good agreement as the three-exponential model is obtained when fitting a bimolecular model, with an added single exponential, to the data for binding of ∆∆-P4 to poly[d(A-T)2] (data not shown):

I(t) ) kB[A]0t/(1 + kB[A]0t) + exp(-kMt) where kB is the bimolecular rate constant, kM the first-order rate constant, and [A]0 the initial concentration of ruthenium complex. Fitting this expression to the data estimates (kB[A]0)-1 to be 391 s and kM-1 to be 8960 s. We notice that the value for the single exponential, kM, as expected, is very similar to that of the slowest process when a fit with three single exponentials is used. To shed further light on the mechanism of threading and the subsequent reorganization of ∆∆-P4, we use the minor groove binding dye DAPI as a probe molecule. Figure 3 shows a titration series where the binding of DAPI to poly[d(A-T)2], when ∆∆-P4 is already bound in its final state, was assessed with circular dichroism (CD) and fluorescence (Figure 3). Since the binding of the ruthenium complex to its final, thermodynamically most favored site is very slow, the samples with poly[d(A-T)2] and ∆∆-P4 were left to equilibrate overnight at 50 °C, whereafter CD and emission spectra were measured with increasing amounts of DAPI. The figure shows the emission from DAPI integrated from 400 to 575 nm (top) and the CDsignal from DAPI integrated from 325 to 410 nm (bottom) plotted against the DAPI concentration at four different ruthenium complex binding ratios. Note a decrease in emission from DAPI at high binding ratios, probably due to self-quenching from more densely bound DAPI molecules. Self-quenching could also be the reason DNA appears fully loaded with DAPI at slightly lower binding ratios when studied with emission than when studied with CD. The fact that the signal from DAPI is smaller both in CD and emission, when the concentration of ∆∆-P4 is increased, indicates, not unexpectedly, that there is

Westerlund et al.

Figure 4. Emission intensity changes for DAPI (monitored at 450 nm, top panel) and ∆∆-P4 (monitored at 630 nm, bottom panel) in the presence of poly[d(A-T)2]. At 300 s, 10 µM DAPI was added to a sample ∆∆-P4 and 80 µM poly[d(A-T)2], which had been mixed at time zero. ∆∆-P4 concentrations increase from 0 to 2, 4, and 6 µM, (in b, 0 µM ∆∆-P4 is excluded). The gray line in the bottom panel corresponds to association of 6 µM ∆∆-P4 to poly[d(A-T)2] without DAPI added. The arrows point in the direction of increasing ruthenium complex concentration. Samples excited at 360 nm. Measurements were done at 50 °C in aqueous buffer (4 mM HEPES, 2 mM MgCl2, 5 mM KCl at pH 7.4).

less space available for DAPI to bind to DNA with an increasing amount of ruthenium complex bound. Comparing the top and bottom panel in Figure 3, it is obvious that there is a larger decrease in the emission signal than in the CD signal. The decreased CD signal with increased ruthenium complex concentration is, most likely, because there are fewer binding sites available for DAPI to bind when the binding ratio for the ruthenium complex is higher, whereas the decreased emission also has a contribution from energy transfer due to closer distances between DAPI and ∆∆-P4. Figure 4 shows the emission intensity from DAPI (top) and ∆∆-P4 (bottom) as a function of time. ∆∆-P4 is added to poly[d(A-T)2] at time zero, whereafter 10 µM DAPI is added at 300 s. The instant increase in emission from ∆∆-P4 seen after 300 s in Figure 4, when DAPI is added, is due to energy transfer from DAPI to the ruthenium complex. The DAPI emission at 300 s is decreased with an increasing concentration of ruthenium complex (Figure 4, top panel) due to more close contacts between DAPI and ∆∆-P4, which leads to more efficient quenching due to energy transfer but also to fewer DAPI molecules bound to DNA. The decay in DAPI fluorescence intensity with time after 300 s may be ascribed to a reorganization of the ruthenium complexes, leading to dissociation of DAPI and/or increased energy transfer from DAPI to ∆∆-P4, due to more close contacts between DAPI and ∆∆-P4. To assess the relative importance of the two effects, the experiment in Figure 4 was repeated, for the highest ruthenium complex concentration (6 µM), but instead monitoring CD (Figure 5). A decrease in CD from DAPI occurrs with approximately the same rate as the emission decrease, consistent with dissociation of DAPI from DNA. Since the decrease in CD is of the same relative magnitude, about 15% of the value in the absence of ruthenium complex, as the decrease in emission, these results suggest that the decrease in emission with time is mainly due to dissociation of DAPI and

Redistribution of Threaded Ruthenium Complex

J. Phys. Chem. B, Vol. 109, No. 44, 2005 21143

Figure 5. Change in integrated CD (full circles) and emission (empty circles, monitored at 450 nm, excitation wavelength was 360 nm) from DAPI. At 0 s, 10 µM DAPI is added to a sample of 6 µM ∆∆-P4 and 80 µM poly[d(A-T)2], which had been mixed 300 s earlier. The values are normalized to the corresponding values for 10 µM DAPI without ∆∆-P4 added. Measurements were done at 50 °C in aqueous buffer (4 mM HEPES, 2 mM MgCl2, 5 mM KCl at pH 7.4).

that the loss of emission due to energy transfer is negligible in this time interval. The faster process after DAPI is added corresponds to an increase in ruthenium complex emission and a decrease in DAPI emission, whereas the slower one corresponds to an increase in emission from both ∆∆-P4 and DAPI. As can be seen after the break in the figure, the emission continues to increase on a very long time scale (over 50 000 s). The curves in Figure 4a,b can after the addition of DAPI (at 300 s) be fitted globally to a biexponential expression with reciprocal rate constants of 900 and 11 000 s. The rate constants are very similar to those of the two slowest processes found in the three exponentials fit in the absence of DAPI, suggesting that the slowest process for DAPI is correlated to the slowest process for ∆∆-P4. It also suggests that the kinetics of the equilibration of the ruthenium complexes is not significantly affected by the presence of the DAPI probe molecule. The slow increase in DAPI emission is not accompanied by any corresponding increase in the CD signal, suggesting that the number of bound DAPI molecules remains the same (Figure 5). A plausible explanation to these findings is that a slow reorganization of ∆∆-P4, accompanied by an increase in emission, is taking place, leading to a decrease in the number of DAPI molecules close enough to a ruthenium complex to be quenched by energy transfer. Since practically all free binding sites of the DNA are occupied, either with ruthenium complex or with DAPI, the simplest explanation for this phenomenon is an increase in the number of ruthenium complexes being neighbors to each other and thus not to DAPI molecules (Figure 6); i.e., the thermodynamically more favorable distribution of ruthenium complexes is characterized by a more cooperative binding than the distribution formed by the initial threading process. Assuming that the only bound DAPI molecules that are quenched are those with at least one ruthenium complex neighbor, we used the generalized McGhee-von Hippel model36 for cooperative binding of multiple ligands to a one-dimensional lattice to model the changes in DAPI CD (assumed to be proportional to the binding density) and emission.37,38 The good fit shows that energy transfer between neighbors only is sufficient to model the data satisfactorily, although we cannot exclude that some energy transfer on longer distances occurs. Taking the binding site coverage parameter n to be 2.5 for DAPI and 4 for ∆∆-P4 and estimating the intrinsic binding constants K to be 1 ×107 M-1 and 2 × 109 M-1, respectively, at total concentrations 40 µM basepairs, 10 µM DAPI, and 6 µM ∆∆-

Figure 6. Schematic picture of the origin of slow rearrangement of ∆∆-P4 on DNA with time. The number of close contacts between ∆∆P4 (yellow) and DAPI (red) is decreased when the ruthenium complexes move toward a more stochastic binding on DNA.

P4, the change in normalized emission and CD in Figure 5 from 2000 to 50 000 s was calculated as a function of the complexcomplex cooperativity parameter. An increase of that parameter from 1 to 2.5 (all other cooperativity parameters assumed to be 1) gave an increase in normalized emission from 0.21 to 0.26 with an almost constant CD of 0.61, in very good agreement with the data depicted in Figure 5. Discussion The strong positive induced CD signal and the bright fluorescence of DAPI when bound in the minor groove of ATrich regions of DNA make this dye a very suitable probe for following binding to DNA. The fact that the binding of ∆∆-P4 to DNA28 is much stronger than for DAPI39 is also important, so the binding of DAPI has minimal effects on the binding of ∆∆-P4. Furthermore, the good spectral overlap of the emission of DAPI with the absorption of ∆∆-P4 allows us to use FRET to rather sensitively follow the effective spatial separation between the ruthenium complex and DAPI on DNA. A timedependent change in emission of DAPI would thus reflect a movement or redistribution of ∆∆-P4, affecting the number of available sites for DAPI to bind to, and also indicate whether the available binding sites are close enough to ∆∆-P4 to be quenched or not. The process after addition of DAPI (t ) 300 s) can be divided into two steps. The first is a relatively fast process where the emission from DAPI decreases with time, whereas the emission from ∆∆-P4 increases. We concluded in Figure 5 that the decrease in emission with time is only due to dissociation of DAPI and that the effect of energy transfer is negligible. Dissociation of DAPI indicates that there are less free binding sites left after the threading process, i.e., that the threaded complexes have a somewhat larger size coverage parameter than the initially bound state, but also agrees with the idea that the complexes bind preferentially avoiding each other during the threading process (Figure 6), since they require large openings on DNA for the threading intercalation to occur.31 This suggests that after an instant, stochastic binding distribution governed mainly by electrostatics, there is an anticooperative effect during

21144 J. Phys. Chem. B, Vol. 109, No. 44, 2005 the threading process, such that the ∆∆-P4-complexes effectively tend to repel each other. The subsequent very slow increase in emission from DAPI, accompanied by an equally slow increase in emission from ∆∆-P4, is not caused by more DAPI molecules binding to DNA, since the CD signal is not affected, but it is rather due to equilibration of ∆∆-P4, away from the anticooperative distribution accompanying the threading, toward a more stochastic binding, so that there are fewer close contacts between DAPI and ∆∆-P4 (Figure 6). In fact, the modeling of the data with the generalized McGhee-von Hippel method36,38 indicates the thermodynamically favored binding mode to be slightly cooperative, with a modest value of 2.5 for the complexcomplex cooperativity parameter. Further indications for the weakly cooperative nature of the redistribution can be seen in Figure 4 (bottom), where the increase in emission from the ruthenium complex at long times is much less pronounced when the ruthenium concentration, and thereby the binding ratio, is lowered three times. This conclusion is further confirmed when studying the DAPI emission (Figure 4, top), where no slow increase in emission is observed at long times for the lowest ruthenium binding ratio. There are two principally different ways by which the rearrangement to a more stochastic binding could occur. Either the complex first dissociates from DNA, possibly by diffusing laterally along the DNA, and binds back in a thermodynamically more favorable binding pocket with a higher emission quantum yield, a kind of shuffling that has been suggested for the sterically similar natural antibiotic nogalamycin.2-4 Alternatively, the complex remains threaded and manages to rearrange itself by moving within DNA, a mechanism that would be anticipated to be extremely slow, as it would require sequential base-pair opening and would thus not allow the complex to diffuse rapidly over large distances. The change in the binding distribution, as judged from the change in the cooperativity parameter, is however rather small, suggesting that the movement within the DNA is in fact a possible mechanism. In conclusion, we have shown that we are able to study the rearrangement of the ruthenium complex on DNA in real time by studying energy transfer from DAPI. The studies show that there is a slow cooperative effect; the ruthenium complexes move from a spread out binding during the intercalation toward the thermodynamically favored distribution. References and Notes (1) Mu¨ller, W.; Crothers, D. M. J. Mol. Biol. 1968, 35, 251-290. (2) Fox, K. R.; Waring, M. J. Biochim. Biophys. Acta 1984, 802 (2), 162-168. (3) Fox, K. R.; Brassett, C.; Waring, M. J. Biochim. Biophys. Acta 1985, 840 (3), 383-392. (4) Fox, K. R.; Waring, M. J. Nucleic Acids Res. 1986, 14 (5), 20012014. (5) Liaw, Y. C.; Gao, Y. G.; Robinson, H.; van der Marel, G. A.; van Boom, J. H.; Wang, A. H. Biochemistry 1989, 28 (26), 9913-9918.

Westerlund et al. (6) Gao, Y. G.; Liaw, Y. C.; Robinson, H.; Wang, A. H. Biochemistry 1990, 29 (45), 10307-16. (7) Egli, M.; Williams, L. D.; Frederick, C. A.; Rich, A. Biochemistry 1991, 30 (5), 1364-1372. (8) Smith, C. K.; Davies, G. J.; Dodson, E. J.; Moore, M. H. Biochemistry 1995, 34 (2), 415-425. (9) Collier, D. A.; Neidle, S.; Brown, J. R. Biochem. Pharmacol. 1984, 33 (18), 2877-2880. (10) Williams, L. D.; Egli, M.; Qi, G.; Bash, P.; van der Marel, G. A.; van Boom, J. H.; Rich, A.; Frederick, C. A. Proc. Natl. Acad. Sci. U.S.A. 1990, 87 (6), 2225-2229. (11) Waring, M. J.; Wakelin, L. P. Nature 1974, 252 (5485), 653-657. (12) Wakelin, S. P.; Waring, M. J. Biochem. J. 1976, 157 (3), 721740. (13) Shafer, R. H.; Waring, M. J. Biopolymers 1980, 19 (2), 431-443. (14) Fox, K. R.; Wakelin, L. P. G.; Waring, M. J. Biochemistry 1981, 20 (20), 5768-5779. (15) O ¨ nfelt, B.; Lincoln, P.; Norde´n, B. J. Am. Chem. Soc. 1999, 121 (46), 10846-10847. (16) O ¨ nfelt, B.; Lincoln, P.; Norde´n, B. J. Am. Chem. Soc. 2001, 123 (16), 3630-3637. (17) Balzani, V.; Ballardini, R. Photochem. Photobiol. 1990, 52 (2), 409-416. (18) Pyle, A. M.; Barton, J. K. Prog. Inorg. Chem. 1990, 38, 413-475. (19) Friedman, A. E.; Kumar, C. V.; Turro, N. J.; Barton, J. K. Nucleic Acids Res. 1991, 19 (10), 2595-2602. (20) Chow, C. S.; Barton, J. K. Methods Enzymol. 1992, 212, 219242. (21) Jenkins, Y.; Friedman, A. E.; Turro, N. J.; Barton, J. K. Biochemistry 1992, 31 (44), 10809-10816. (22) Hiort, C.; Lincoln, P.; Norde´n, B. J. Am. Chem. Soc. 1993, 115 (9), 3448-3454. (23) Haq, I.; Lincoln, P.; Suh, D.; Norden, B.; Chowdhry, B. Z.; Chaires, J. B. J. Am. Chem. Soc. 1995, 117 (17), 4788-4796. (24) Norde´n, B.; Lincoln, P.; Åkerman, B.; Tuite, E. Met. Ions Biol. Syst. 1996, 33, 177-252. (25) Lincoln, P.; Broo, A.; Norde´n, B. J. Am. Chem. Soc. 1996, 118 (11), 2644-2653. (26) Erkkila, K. E.; Odom, D. T.; Barton, J. K. Chem. ReV. (Washington, DC) 1999, 99 (9), 2777-2795. (27) Friedman, A. E.; Chambron, J. C.; Sauvage, J. P.; Turro, N. J.; Barton, J. K. J. Am. Chem. Soc. 1990, 112 (12), 4960-4962. (28) Lincoln, P.; Norde´n, B. Chem. Commun. 1996, 18, 2145-2146. (29) Wilhelmsson, L. M.; Westerlund, F.; Lincoln, P.; Norde´n, B. J. Am. Chem. Soc. 2002, 124 (41), 12092-12093. (30) Wilhelmsson, L. M.; Esbjo¨rner, E. K.; Westerlund, F.; Norde´n, B.; Lincoln, P. J. Phys. Chem. B 2003, 107 (42), 11784-11793. (31) Nordell, P.; Lincoln, P. J. Am. Chem. Soc. 2005, 127 (27), 96709671. (32) Yun, B. H.; Kim, J.-O.; Lee, B. W.; Lincoln, P.; Norden, B.; Kim, J.-M.; Kim, S. K. J. Phys. Chem. B 2003, 107 (36), 9858-9864. (33) Lee, B. W.; Moon, S. J.; Youn, M. R.; Kim, J. H.; Jang, H. G.; Kim, S. K. Biophys. J. 2003, 85 (6), 3865-3871. (34) Hansen, J. B.; Koch, T.; Buchardt, O.; Nielsen, P. E.; Norden, B.; Wirth, M. J. Chem. Soc., Chem. Commun. 1984, 8, 509-511. (35) Wirth, M.; Buchardt, O.; Koch, T.; Nielsen, P. E.; Norden, B. J. Am. Chem. Soc. 1988, 110 (3), 932-939. (36) McGhee, J. D.; von Hippel, P. H. J. Mol. Biol. 1974, 86 (2), 46989. (37) Lincoln, P.; Tuite, E.; Norden, B. J. Am. Chem. Soc. 1997, 119 (6), 1454-1455. (38) Lincoln, P. Chem. Phys. Lett. 1998, 288 (5, 6), 647-656. (39) Kapuscinski, J.; Skoczylas, B. Nucleic Acids Res. 1978, 5 (10), 3775-3799.