same direction in the intercalation complex and they overlap the G C base pair in the binding site. The
nogalamycin–DNA contacts that give rise to specificity occur mainly in the major groove. This is in contrast
to the daunomycin–DNA complex where interactions are predominantly in the minor groove. The uncharged
nogalamycin sugar that occupies the minor groove appears to interact only very weakly with the DNA. In
a way similar to daunomycin, the long axis of nogalamycin is perpendicular to the long axis of the base
pairs. This arrangement leads to a slight buckling of base pairs that constitute the intercalation site, such
that they wrap around the bound drug. This results in improved van der Waals contacts. As expected for a
threading intercalator, the mechanism of nogalamycin binding to duplex DNA is not simple. For intercal-
ation to occur, one or other of the attached sugar moieties has to thread through the intercalation site.
These structural constraints lead to slow association and dissociation kinetics.
The general features of the threading mode have been illustrated by stopped-flow kineticsmeasure-
ments conducted by Wilson and co-workers on two groups of disubstituted anthracene-9,10-diones
(anthraquinones)synthesised in the laboratories of Jenkins and Neidle (Figure 9.12).^27
Several such anthraquinone compounds have been synthesized and they have been shown to bind to
duplex, triplex, and quadruplex DNA by intercalation. The 1,4-bis(amino) functionalized compound, mitox-
antrone, was in clinical use for the treatment of breast cancer. Wilson, Jenkins and Neidle have systemat-
ically examined the effects of alteration of substituents on the anthraquinone ring as well as alteration of
substituent positions. From molecular modelling studies, it was predicted that anthraquinones substituted
at either the 1,4- or 1,8-positions with cationic side chains should intercalate into DNA with both side chains
binding in the same groove through classical intercalation. However, drugs substituted at either the 1,5- or
2,6-positions on the anthraquinone ring should intercalate through a threading mode with one side chain
occupying each groove.
DNA binding measurements, conducted as a function of salt concentration and temperature, concur with
the predictions of molecular modelling. 1,4- or 1,8-difunctionalized anthraquinones have association rate
constants of about 2 106 M^1 s^1 , while for 1,5- or 2,6-compounds these values are 2 105 M^1 s^1.
Hence the compounds that bind by threading have about 10-fold slower association kinetics than do classi-
cal intercalators. The dissociation rate constants are approximately 10 s^1 or greater for classical intercalators
and 5 6s^1 for threading intercalators. Interestingly, compensatory effects in association and dissoci-
ation rate constants give rise to equilibrium binding constants that are closer than kinetic constants for
classical versusthreading intercalation. The slower observed kinetics for the threading mode arise because
360 Chapter 9
Figure 9.12 Functionalised anthracene-9,10-diones showing substituents at the 1,4- and 2,6-ring positions