and H17.6 kJ mol^1 at 30.1°C). The binding-induced change in heat capacity ( Cp) is therefore
estimated as 1.38 kJ mol^1 K^1. This gives an overall thermodynamic profile of positive enthalpy and
entropy together with a negative heat capacity change, which is generally characteristic of hydrophobic
interactions and correlates well with X-ray crystal structures of this complex. The binding reaction is entrop-
ically driven and the favourable entropy term arises from the release of structured water and cations from
the minor groove and/or the ligand into bulk solvent upon binding. Another origin for the positive entropy
is the release of condensed counter-ions from the DNA helix upon binding of the cationic ligand.
The heat capacity change can be related to changes in the surface area exposure (Section 9.6.4 and
Equation 9.16). Binding of Hoechst 33258 to A3T3 results in a 20% loss of solvent-accessible surface
relative to the individual components. The majority of surface removed from exposure to solvent upon
binding is nonpolar rather than polar. Application of Equation 9.16 to the Hoechst 33258–A3T3 system
gives a calculated Cpof 1.37 kJ mol^1 K^1 which is in good agreement with the calorimetrically deter-
mined heat capacity change of 1.38 kJ mol^1 K^1.
The empirical relationship used to calculate Cpwas originally based on the heats of solvent transfer of
small model compounds such as amino acids, amides, and hydrocarbons. Later it was extended and applied
to protein folding–unfolding equilibria and to protein–DNA interactions.^50 The Hoechst 33258–A3T3
calorimetric study was the first demonstration that the same relationship is valid for low molecular weight
DNA binding drugs. The values used in Equation 9.16 are derived only on data from drug–DNA inter-
actions as it is now established that change in surface area exposed to solvent on ligand binding provides
a tentative link between structural and thermodynamic data.
There have been several reports of calorimetrically measured heat capacity changes for both groove
binders and intercalators. In most cases there is good agreement with theoretical predictions of heat cap-
acity change derived from predictive algorithms of the type shown in Equation 9.16. Figure 9.17 shows a
good correlation of calculated versus measured heat capacity change for a variety of systems.
A detailed partitioning of Gobsinto its component parts gives considerable insight into the forces
responsible for the binding process (Section 9.6.4 and Equation 9.14). In the Hoechst 33258–A3T3 system,
the contribution of each of the five terms in Equation 9.14 to Gobshas been estimated. The variation of
Kobswith salt concentration and application of polyelectrolyte theory revealed that Gpeis 7.36 kJ mol^1.
Reversible Small Molecule–Nucleic Acid Interactions 369
Figure 9.17 The relationship between calorimetrically measured heat capacity change and the heat capacity
change calculated from changes in solvent accessible surface area for twelve different drug–DNA
systems (red: intercalators, black: groove binders). The line is a linear fit to the data (r0.97)