1710
J. Phys. Chem. B 2009, 113, 1710–1721
Rational Design of Anthracene-Based DNA Binders Michael R. Duff, Jr., Vamsi K. Mudhivarthi, and Challa V. Kumar* Department of Chemistry, U-3060, 55 North EagleVille Road, UniVersity of Connecticut, Storrs, Connecticut 06269-3060 ReceiVed: August 11, 2008; ReVised Manuscript ReceiVed: October 16, 2008
Achieving the goal of rational design of DNA-binding ligands is important, and many inroads have been made in this direction. Toward that goal, we report a simple, systematic, and quantitative approach to design DNA-binding anthracene derivatives. Current data show that the binding free energies (∆G°) as well as enthalpies (∆H°) are related to specific structural features of the binders. Systematic design of anthracene probes, for example, indicated that the affinity can be enhanced via the introduction of methylene groups. Each methylene group contributed, on an average, -0.08 ( 0.002 kcal/mol (at 1 M ionic strength, 293 K) toward the total binding free energy. Binding of the probes to DNA depended on ionic strength, and ionic strength studies were used to factor out to parse free-energy contributions due to specific interactions. The intrinsic free-energy contributions (∆GMol) of the probes are obtained by factoring out contributions from ionic interactions, hydration, conformational changes, polyelectrolyte effect, and the loss of rotational/ translational motion. A strong, linear correlation was noted between ∆GMol and the number of methylene groups present in the probe, and the correlation indicated free-energy contributions of -1.49 kcal/mol per methylene (at 50 mM NaCl, 293 K). This important observation provides a convenient handle to systematically fine-tune the intrinsic affinities of DNA binders. ∆H values also showed clear trends, and each methylene contributed +0.28 kcal/mol toward the overall binding enthalpy (at 50 mM NaCl, 293 K), and this aspect is useful to fine-tune ∆H contributions to binding. These important physical insights, derived from systematic modifications of the side chains of the DNA binders, are useful in the rational design of novel DNA binders. Introduction Developing a quantitative understanding of the interactions of small molecules with DNA is helpful for the rational design of DNA-binding ligands for various applications.1 Even though such interactions are numerous and often interdependent, parsing the free-energy contributions of these interactions, in specific cases, resulted in enormous progress in this area.2 To date, freeenergy contributions of specific structural elements of the ligands toward the binding are poorly identified, and progress in this area is critical for the rational design of DNA binders. For example, what are the contributions of hydroxyl, amino, ester, amide, or CH2 groups to the enthalpies and free energies of binding, in a given class of compounds? Are these values obtained for one class of ligands relevant to other classes of DNA binders? This important issue of structural thermodynamics is not well addressed in the literature, despite the fact that several authors have identified this gap in our understanding of DNA-ligand interactions.3
∆Gobs ) ∆Gconf + ∆Gr+t + ∆Grot + ∆Ghyd + ∆GPE + ∆GMol (1) For example, the overall binding free energy (∆G°) of a ligand to DNA is a sum of several positive and negative terms assigned to specific interactions (eq 1),2 where ∆Gconf is the free-energy change due to conformational changes of the DNA as well as the small molecule upon the DNA-ligand complex formation, ∆Gr+t is the free-energy change due to the loss of * To whom correspondence
[email protected].
should
be
addressed.
E-mail:
the rotational and translational motions when the bimolecular complex is formed, ∆Grot is due to restrictions on the rotations around specific bonds of the ligand in the DNA complex, ∆Ghyd is the contribution due to hydration/dehydration accompanying the binding event, ∆GPE is the contribution due to the polyelectrolyte effect, and ∆GMol is assigned to the free-energy change assigned to specific interactions such as H-bonding, van der Waals interactions, electrostatic interactions, and others. Systematic accounting of the individual free-energy terms in eq 1 would provide information regarding the magnitude of ∆GMol, which is not well established for most molecules. Furthermore, by systematically varying the structural features of the ligand and determining ∆GMol, one could obtain the freeenergy contributions of specific structural elements of the ligand to the overall binding. This information is currently unavailable, but it would facilitate the rational design of DNA binders. Here, we report a quantitative, general approach to evaluate the enthalpy and free-energy contributions of CH2 groups to the binding of anthracene ligands to DNA, a step toward the rational design of DNA binders. Anthracene derivatives are chosen for the current studies because they received considerable attention as potential antitumor drugs and they offer a convenient class of DNA binders for rational design.4 Pseudourea is an example of anthracenes tested in clinical trials, but it was soon withdrawn due to various reasons, and hence, swift progress in the rational design of novel anthracene ligands is important.5 Our long-term interest has been to make quantitative evaluations of the contributions of specific structural features of the anthracene ligands for binding to DNA. Our approach has been to systematically vary the substituents at the 9 and 10 positions of the anthracene ring system and parse the contributions of
10.1021/jp807164f CCC: $40.75 2009 American Chemical Society Published on Web 01/21/2009
Rational Design of Anthracene-Based DNA Binders CHART 1: Two Major DNA Binding Modes of Groove Binding (Netropsin, PDB 101D) and Intercalation (Benzo[a]anthracene, PDB 1DL4) Ligandsa
J. Phys. Chem. B, Vol. 113, No. 6, 2009 1711 Calf thymus DNA (CT DNA) is chosen for these investigations for numerous reasons. The sequence heterogeneity of CT DNA represents the genomic DNA of the living cell. Therefore, binding studies with this DNA are relevant to the in vivo behavior of the anthracene probes. A large number of studies on anthracene probes are already reported with this type of DNA,7–10,12 and these data are useful for comparisons with the current findings. Our data show significant free-energy contributions of the long side chains of anthracene derivatives, in calorimetric and spectroscopic investigations. Experimental Section
a The DNA-bound ligands are shown in blue, which are taken from the high-resolution crystal structures from the literature.6
CHART 2: Structures of Anthracene Derivatives Prepared for This Study
specific structural elements of the probe to specific attributes of the DNA-ligand complex. Such properties include, but are not limited to, the overall binding affinities, enthalpies of binding, the dominant mode of binding, on- and off-rates from the helix, and sequence selectivity for DNA binding. Here, we present some of the first reports of the intrinsic enthalpy and free-energy contributions of methylene groups toward DNA binding of ligands, and the current approach may be extended to other classes of DNA ligands. Progress in this area will aid in the rational design of novel DNA-binding ligands whose properties can be tailored to meet desired specifications. Anthracene derivatives shown in Chart 2 are chosen for the current studies due to numerous advantages of this class of ligands which include favorable absorption and emission characteristics and facile synthetic routes to make a variety of anthracene derivatives, and there exists substantial literature supporting their binding to DNA by specific binding modes.4c,5a,7–9 The binding affinities of the anthracene probes increased with an increase in the number of hydroxyethyl groups,10 and the octyl amine side chain increased the binding affinity by 70fold.11 These observations challenged us to test if anthracene derivatives can be rationally designed with predictable DNA binding affinities and thermodynamic properties. One approach is to quantitate the DNA binding affinities of closely related anthryl probes where specific structural changes have been introduced in the ligand. Such measurements will facilitate the contributions of selected structural features to specific interactions. Our hypothesis is that the ∆GMol increases linearly with the number of methylene groups present in the probe, in a given class of molecules, due to increased hydrophobic interactions. As shown in Chart 2, the number of methylene groups is increased systematically from 2 to 26 (Chart 2), with the 9,10bis(aminomethyl)anthracene unit as the common motif. From quantitative binding data, we deduced the average contributions of each methylene group to the overall binding free energy with DNA.
Calf thymus DNA (CT DNA, type I) was purchased from Sigma Chemical Co., and the sample was purified according to published protocols.13 Concentrations of CT DNA solutions were determined by absorption spectroscopy by using the extinction of 13 600 M-1 cm-1 at 260 nm.14 The purity of the CT DNA was checked by monitoring the ratio of absorbance at 260 nm to that at 280 nm (1.8). Stock solutions of DNA were prepared with Tris buffer (5 mM Tris, 50 mM NaCl, pH 7.2), and the DNA concentration is expressed in terms of DNA base pairs, unless stated otherwise. The DNA binding properties, in representative cases, were also examined using sonicated CT DNA, and these parameters are essentially the same as those obtained with type I DNA. BAMAC, BEDA, and BPA were synthesized according to literature procedures.5a,8,12 The NMR spectra, the absorption spectra, and the melting points of these substances matched with those reported.15 The analytical data on BEDA and BPA have been published earlier.9,12 BAMAC data: 54% yield; mp 238 °C; 1HNMR (D2O) (ppm, multiplicity, relative intensity) 8.43, m, 4; 7.86, m, 4; 5.21, s, 4; and absorption maxima: 352, 370, 390 nm. BEDA data: 40% yield; mp 185-186 °C; 1HNMR (D2O) (ppm, multiplicity, relative intensity) 8.43, multiplet, 4; 7.76, multiplet, 4; 5.21, broad, 4; 3.52, multiplet, 4; 3.21, multiplet, 4.; absorption maxima: 355, 374, 394 nm. Synthesis of 9,10-Bis(4-N′,N′-(1-aminobutyl)aminomethyl)anthracene (BBDA). To 1,4-diaminobutane (2 mL, 18 mmol, 13 equiv) in 50 mL of acetonitrile was added a solution of 9,10bis(bromomethyl)anthracene (0.5 g, 1.4 mmol, in 25 mL of acetonitrile) over an hour. The reaction was allowed to proceed overnight. The product mixture was concentrated under vacuum and then washed with hexane (3×, 20 mL) and ether (3×, 20 mL). The resulting solid was dissolved in methylene chloride and washed three times with 0.1 N HCl. The aqueous layer was then carefully made basic to pH 9.5 and then extracted four times with methylene chloride. The organic layer was dried, the resulting residue was then dissolved in a minimal amount of 0.1 N HCl, and the product was crystallized with the addition of 10-fold (v/v) isopropanol. This gave a pure sample (27% yield): mp 180 °C (decomp), 1HNMR (the base form in CDCl3) (ppm, multiplicity, relative intensity) 8.39, m, 4; 7.54, m, 4; 4.73, s, 4; 2.92, t, 4; 2.73, t, 4; 1.68, m, 4; 1.54, m, 4; 1.31, m, 4; and absorption maxima: 354, 373, 394 nm. Exact mass gave m/z ) 379.5709; the expected mass was 379.5689, which confirmed the structure. Synthesis of 9,10-Bis(8-N′,N′-(1-aminooctyl)aminomethyl)anthracene (BODA). To 1,8-diaminooctane (2.98 g, 21 mmol, 15 equiv) in 150 mL of toluene was added a solution of 9,10bis(bromomethyl)anthracene (0.5 g, 1.4 mmol, in 50 mL of acetonitrile) over an hour. The reaction was allowed to proceed overnight, 50 mL of acetonitrile added to the reaction mixture, and the reaction continued for an additional 24 h. The product mixture was concentrated under vacuum and then washed with
1712 J. Phys. Chem. B, Vol. 113, No. 6, 2009
Duff et al.
hexane (3×, 20 mL) and ether (3×, 20 mL). The resulting solid was repeatedly dissolved in methylene chloride, and unreacted diamine was filtered until no more precipitate formed. The sample was crystallized from a 1:5 mixture of methylene chloride and hexane. This gave a pure sample (37% yield): mp 200 °C (decomp). 1HNMR (DMSO-d/10% D2O) (ppm, multiplicity, relative intensity) 8.57, m, 4; 7.82, m, 4; 5.29, s, 4; 3.27, t, 4; 2.78, t, 4; 1.76, m, 4; 1.55, m, 4; 1.31, m, 16; and absorption maxima: 354, 373, 393 nm. Exact mass gave m/z ) 491.7868; 491.7837 was the expected mass, which confirmed the structure. Synthesis of 9,10-Bis(12-N′,N′-(1-aminododecyl)aminomethyl)anthracene (BDDA). To 1,12-diaminododecane (2 g, 10 mmol, 7 equiv) in 150 mL of toluene was added a solution of 9,10-bis(bromomethyl)anthracene (0.5 g, 1.4 mmol, in 50 mL of acetonitrile) over an hour. The reaction was allowed to proceed overnight, 50 mL of acetonitrile added to the reaction mixture, and reaction continued for an additional 24 h. The product mixture was concentrated under vacuum and then washed with hexane (3×, 20 mL) and ether (3×, 20 mL). The resulting solid was dissolved in methylene chloride, and unreacted diamine was filtered. The product and unreacted amine were extracted from the methylene chloride solution by 0.1 N HCl (3×, 30 mL). The water extract was made basic to pH 9.5, and the product was extracted with methylene chloride. This gave a pure sample (28% yield): mp 230 °C (decomp). 1HNMR (in CD3OD) (ppm, multiplicity, relative intensity) 8.37, m, 4; 7.55, m, 4; 4.70, s, 4; 2.76, m, 4; 2.58, m, 4; 1.58, m, 4; 1.39, m, 4; 1.26, m, 32; and absorption maxima: 354, 373, 393 nm. Exact mass gave m/z MH+ ) 603.5379 and an elemental composition of C40H67N4, which confirmed the structure. Isothermal Titration Calorimetry. Isothermal titration calorimetry studies were carried out on a VP-ITC from MicroCal Inc. to estimate the binding parameters by following a method reported earlier.10 A brief procedure is given here. CT DNA (100 µM, 5 mM Tris, 50 mM NaCl, pH 7.2) was loaded into the calorimetric cell (1.4167 mL), and the probe solution (200-600 µM probe, 5 mM Tris, 50 mM NaCl, pH 7.2) was loaded into the syringe. Heat produced or absorbed due to each addition of the titrant was recorded, and the corresponding thermograms were analyzed using Origin software (v.5.0, Microcal Inc.). The binding constant, binding site size, enthalpy of binding, entropy of binding, and the free energy of binding were estimated in triplicate measurements. The heat released or absorbed (Q) during the titration was related to the molar heat of ligand binding (∆H), the volume of the sample cell (Vo), the bulk concentration of the ligand (Xt), the concentration of DNA (Mt), the intrinsic binding constant (Kb), and the number of binding sites (n) by the following eq 216
Q)
[
nMt∆HVo Xt 1 1+ + 2 nMt nKbMt
(
1+
Xt 1 + nMt nKbMt
)
2
-
]
4Xt (2) nMt
Binding Enthalpy Measurements. These were carried out in a model-independent manner using ITC.17 CT DNA (300 µM, 5 mM Tris, 50 mM NaCl, pH 7.2) was loaded into the cell, and the BDDA solution (600 µM, 5 mM Tris, 50 mM NaCl, pH 7.2) was loaded into the syringe. A small volume of BDDA (3 µL) was injected during six injections to ensure that all injected BDDA has been bound to CT DNA at each injection. Heat produced or absorbed due to each addition of the titrant was
recorded, and the binding enthalpy was calculated using Origin software (v.5.0, Microcal Inc.). Absorption Spectral Studies. Absorption spectra were recorded on an Ocean Optics UV3000 spectrophotometer, which has been interfaced with an Apple iBook computer. In the absorption titrations, the concentration of the probe was kept constant while increasing the concentration of CT DNA. The data have been corrected for small increases in the sample volume due to each addition of the DNA solution. Binding Plots. The spectral changes observed in the absorption measurements were used to calculate the intrinsic binding constant (Kb) using both the Scatchard equation and the neighbor exclusion model (eq 3).18,19 In eq 3, Cf is the concentration of the free probe, r ) Cb/[DNA], where Cb is the concentration of the bound probe, and N is the binding site size in base pairs (1/n). The concentration of the bound probe is given by Cb ) Ai - εfCt/∆ε, where Ct is the total concentration of the probe, Ai is the absorbance after the ith injection of DNA, and ∆ε ) εf - εb. Note that εf and εb are the extinction coefficients of the free and bound probes, respectively. Half-reciprocal plots of absorbance versus 1/[DNA] were used to obtain εb, and the binding plots were constructed. The percent of hypochromism was calculated as equal to (εf - εb)/100.
r (1 - nr) ) Kb(1 - nr) Cf [1 - (n - 1)r]
[
n-1
]
(3)
Ionic Strength Studies. The intrinsic binding constants were determined as a function of ionic strength to factor out the electrostatic contributions by polyelectrolyte theories.20 Briefly, buffers with 5 mM TrisHCl, pH 7.2, were prepared with increasing NaCl concentrations (10-200 mM), and the binding constants were determined by the absorption spectral method, as stated above. According to the polyelectrolyte theory,21 the natural log of the binding constant was plotted against the natural log of the molar ionic strength, and the effective charge on the probe was calculated according to eq 4
S)
∂ ln(Kb) ∂ ln[Na+]
) -2N(Ψ - Ψ*) - ZΨ*
(4)
where S is the slope of the plot, N is the binding site size in base pairs, Ψ is the fraction of counterions associated per DNA-phosphate (0.88),22 Z is the number of ions released from the helix per probe molecule bound, and Ψ* is the fraction of counterions associated per DNA-phosphate in the intercalated form (0.82).21 The binding constant at 1 M ionic strength was obtained from the intercept of the plot of the natural log of the binding constant versus the natural log of the molar ionic strength. The binding of BEDA at 1 M ionic strength has been calculated from literature reports.23 Circular Dichroism (CD) Studies. The CD spectra were recorded on a JASCO J-710 spectropolarimeter interfaced with a Dell Optiplex personal computer by using software from JASCO. Solutions containing the probe and CT DNA were placed in a quartz cell (1 cm path length), and the spectra were recorded in the 300-500 nm region, while a shorter path length cuvette (0.2 cm) was used to record the spectra in the 200-300 nm region. The operating parameters of 1 nm bandwidth, sensitivity of 5 or 20 millidegrees and response time of 4 s, were used to average up to 16 scans for each sample. Ethidium Displacement Studies. Ethidium displacement studies were performed on a home-built fluorescence spectrom-
Rational Design of Anthracene-Based DNA Binders
J. Phys. Chem. B, Vol. 113, No. 6, 2009 1713
Figure 1. The raw ITC data (A) and the thermogram (B) of BAMAC (400 µM) binding to CT DNA (100 µM) in 5 mM TrisHCl (50 mM NaCl pH 7.2, 298 K). The solid line corresponds to the best fit to the data, with parameters n ) 0.37 ( 0.02, K ) 2.1 ( 0.5 × 105 M-1, ∆H ) -1.6 ( 0.1 kcal/mol and ∆S ) 18.8 ( 1.2 cal/mol/K.
TABLE 1: Thermodynamic Parameters for the Binding of Anthryl Probes to CT DNA from ITC Data (Tris buffer, 50 mM NaCl, 298 K) ligand
1/n
Kb (M-1)
∆H (kcal/mol)
∆S (cal/molK)
∆G (kcal/mol)
BAMAC BEDA9 BBDA BPA BODA BDDA
2.7 ( 0.1 4.5 ( 1.6
2.1 ( 0.5*105 1.4 ( 0.5*105
18.8 ( 0.7 11.6 ( 1.6
-7.2 ( 0.2 -7.0 ( 0.6
9.3 ( 1.7
6.5 ( 1.1*105
-1.6 ( 0.1 -3.5 ( 0.4 -1.5 ( 0.2 -1.9 ( 0.2 0 +2.7 ( 1.9
20 ( 0.2
-7.9 ( 0.2
eter using SLM Aminco 4800 optics and controlled with a Chipmunk Basic program developed in-house for an Apple Macitosh G3 computer. Spectra were recorded by exciting at 546 nm at 1 nm resolution. A typical experiment consisted of dissolving ethidium bromide (5 µM) in buffer (5 mM TrisHCl, 100 mM NaCl) and adding CT DNA to bind ethidium, with a little excess ethidium, and recording the spectra. Aliquots of anthryl probes were added to the ethidium-DNA solution and allowed to equilibrate for at least 5 min prior to recording the spectra. DSC Measurements. Thermal denaturation experiments were performed on a Calorimetry Sciences Corporation (CSC) 6100 Nano II differential scanning calorimeter (DSC), as described earlier.24 Typically, DNA solutions or DNA-ligand solutions (5 mM citrate, 0.44 mM NaCl, pH 7.2) were scanned against the buffer from 25 to 120 °C, at a heating rate of 2 °C/min. The buffer itself did not produce any endothermic or exothermic transitions in this temperature range. Reversibility of the thermal transition was examined by recording the DSC trace during the second heating cycle. The peak transition temperature (Tm), the full width at half-maximum (fwhm), and the observed area under the transition (integral Cp dT ) ∆H) are modelindependent.25 The temperature coefficient of the pKa of Tris buffer is -0.027 per °C,26 while that of citrate is near 0.27 Therefore, citrate buffer was used to minimize the temperature dependence of the pKa, but no significant differences have been noted between the measurements with these two buffers. Hydrophobic Area Calculations. Theoretical calculations of the surface areas for the anthryl probes were calculated using Spartan06 for Macintosh (Wavefunction, Inc.; Irvine, CA) on an Intel Core Duo iMac. Hydrophobic surface areas were determined from the difference in total surface area and polar surface area for each probe.
Results Calorimetric and spectroscopic studies of anthracene derivatives (Chart 2) show that the long side chains exert a significant influence on DNA binding by these probes. Binding enthalpies and free-energy terms are obtained from the ionic strength studies, and a simple strategy to control DNA binding affinity is deduced from these data. Titration Calorimetry. Each addition of a solution of BAMAC (400 µM, Tris buffer) to a solution of CT DNA (100 µM, Tris buffer) resulted in the release of heat until the binding saturated (Figure 1A). The time delay between the injections, the volume of the injectant, and the number of injections were adjusted such that the binding completed during the titration. In separate experiments, the heats of dilution of BAMAC (upper curve, offset for clarity, Figure 1A) and CT DNA (not shown) have been measured. The area under each peak of the above plots was integrated, the corresponding heats of dilution was subtracted, and the thermogram was constructed (Figure 1B). The ITC data indicated that the binding of BAMAC to CT DNA was exothermic, and binding saturated toward the end of the titration. The above thermogram (Figure 1B) was fitted to eq 2, starting with different initial guess values, and the best fit to the data required a single-site binding model. Data from multiple titrations were analyzed and averaged over several measurements, and these indicated the binding parameters 1/n ) 2.7 ( 0.1 base pairs per anthracene, Kb ) 2.1 ( 0.5 × 105 M-1, ∆H ) -1.6 ( 0.1 kcal/mol, and ∆S ) 18.8 ( 0.7 cal/ mol K (Table 1). Calorimetric titrations with other anthracene derivatives were also carried out, and the data were analyzed as appropriate. Binding parameters of BEDA and BPA were previously reported by this laboratory,8,12 and both indicated exothermic binding, similar to BAMAC (Table 1). Binding enthalpies of BBDA, BODA, and BDDA with CT DNA have been measured by
1714 J. Phys. Chem. B, Vol. 113, No. 6, 2009
Duff et al.
Figure 2. (A) Absorption spectra of BAMAC (20 µM) in the presence of increasing concentrations of CT DNA (0-225 µM). (B) Absorbance plot of BODA (11 µM) binding to CT DNA (0-87 µM) in 5 mM TrisHCl and 100 mM NaCl (pH 7.2). (C) Absorbance plot of BDDA (18 µM) binding to CT DNA (0-70 µM). All titrations were carried out in 5 mM TrisHCl and 100 mM NaCl (pH 7.2) in 1 cm cuvettes at 293 K.
TABLE 2: Spectroscopic Properties of the Probes Bound to CT DNA (Tris buffer, 50 mM NaCl, pH 7.2, at 293 K)a probe
isosbestic points (nm)
λmax free (nm)
λmax bound (nm)
∆E (cm-1)
BAMAC BEDA BBDA BPA BODA BDDA
397, 377 399, 384 400 402, 386 399 402b
390 394 393 397 393 394
400 403 403 405 402 402
641 549 549 498 570 505
a
% hypochromism (at nm)
εfree (M-1 cm-1) (at nm)
εbound (M-1 cm-1) (at nm)
67% (370) 70% (394) 71% (374) 67% (397) 65% (373) 65% (374)
6100 (370) 10200 (394) 7160 (374) 16000 (397) 5800 (373) 6200 (374)
2000 (370) 2925 (394) 2100 (374) 4800 (397) 2028 (373) 2200 (374)
Typically, errors are less than (1%. b From data at 10 mM NaCl
Figure 3. Neighbor-exclusion model of Scatchard plots for BAMAC (thick line), BBDA (thin line), BODA (dashed line), or BDDA (dash-dot line) with CT DNA (5 mM Tris, 100 mM NaCl, pH 7.2, at 293 K).
titrating small amounts into the sample cell, in a modelindependent manner.28 The concentrations of the ligand were adjusted, and excess DNA was used so that the ligand added in each aliquot was nearly completely bound to DNA, and thus, the integrated area under each peak indicated the enthalpy of
binding for each addition. These indicated exothermic binding of BBDA with an ∆H of -1.5 kcal/mol, while the binding of BODA to CT DNA showed only weak enthalpy that was no greater than that of the dilution of BODA (data not shown). Therefore, ∆H for BODA binding is nearly equal to zero, and the titration of BDDA into CT DNA showed endothermic binding with an ∆H of +2.7 ( 1.9 (Table 1). The lack of a favorable enthalpic contribution to the binding of BODA or BDDA suggests that the binding of these two ligands is entirely entropy-driven. Absorption Spectral Studies. The absorption spectra of BAMAC (20 µM probe, 5 mM Tris, 100 mM NaCl, pH 7.2) recorded in the presence of increasing concentrations of CT DNA (0-225 µM) are shown in Figure 2A. The DNA concentration was increased in small increments until no further changes were apparent in the absorption spectra, and the spectra were corrected for small volume changes during the titration. The data indicate a strong decrease in the absorption at the peak positions, considerable broadening of the vibronic bands, and the red shift of the peaks by 9-10 nm (Table 2). These changes are comparable to, or larger than, those reported with many other anthryl probes.8–12,29 Also note that red shifts were not observed when the probe solutions were titrated with single-stranded DNA
TABLE 3: Binding Parameters Estimated from the Absorption Titrations at 293 K property/probe f
BAMAC
BEDA
BBDA
BPA
BODA
BDDA
# of methylenes # of amines # of charges at pH 7 Kb/µM (50 mM NaCl) neighbor exlusion model Kb/µM (50 mM NaCl) Scatchard model Kb/µM (100 mM NaCl) neighbor exlusion model Kb/µM (100 mM NaCl) Scatchard model
2 2
