125
• correlation analysis of the "structure-activity" and selecting the “compound hits”
for directed (purposeful) chemical optimization based on the results of biochemical assays;
• chemical synthesis and optimization of new inhibitors based on the analysis of
“structure–activity”, selectivity and computer simulation data.
From this viewpoint the clear understanding of structural aspects of protein–ligand
interaction is a key factor for the correct prediction of inhibitor affinity. Unfortunately, structural information about binding sites for non-nucleoside inhibitors on
the surface of DNA-dependent DNA polymerases is rather limited (see the previous
section).
At the same time, researchers start to use various computational approaches in
the rational design procedure of new (nucleoside analogs or nucleoside containing)
inhibitors of DNA-dependent DNA polymerases (especially, human pols). Spectrum of applicable approaches is sufficiently broad and includes all the above-mentioned methods of the structural analysis in silico.
Richartz and co-workers [157] have successfully applied the methods of molecular docking and molecular dynamics for the analysis of structural action mechanisms
of several nucleotide analogs—potential inhibitors of human DNA polymerase α
(pol α), and of the well-known non-nucleotide pol α inhibitor aphidicolin [56, 157].
In molecular dynamics simulations, aphidicolin occupied the catalytic centre,
but acted in a not truly competitive manner with respect to nucleotides. It destabilized the replicating “closed” form of the pol alpha and transferred the enzyme
into the inactive “open” conformation [157]. This result is consistent with recent
experiments on the binding mode of aphidicolin. Unfortunately, aphidicolin could
not be introduced into therapy because of its toxicity and rapid metabolism after
systemic application [158]. Among studied ‘nucleotides’, the highest potential for
selective pol α inhibition was established for 2-butylanilino-dATP (BuAdATP). The
butylphenyl moiety of BuAdATP occupies a lipophilic pocket, formed by the residues Leu960, Leu972, Val976, Ile869, Tyr865 and Tyr957. These lipophilic interactions, coupled with hydrogen bonds between BuAdATP, template nucleotide and
side chains of residues Tyr865 and Lys950, are likely to be responsible for the good
inhibition efficiency of the butylanilino derivatives. The lowest abilities to inhibit
human pol α were demonstrated for lamivudine-TP and zidovudine-TP (both compounds in three phosphate form) [157].
Later Höltje and co-workers (from the same research team) reported the development of several new human polymerase α inhibitors applicable for skin tumor
treatment (in order to design new drugs for actinic keratosis and squamous cell
carcinoma) [159] (Fig. 4.10). To study the binding modes of these compounds, the
same computational approaches as described in Richartz et al. [157] were used.
It was shown that the compound HM1-TP forms two hydrogen bonds with the
DNA template nucleotide, two hydrogen bonds with the side chain of Lys950 and
one hydrogen bond each with the backbone of Tyr865 and the side chain of Arg922
in the active center of human pol α. At the same time BuP-OH-TP forms two hydrogen bonds with the DNA template nucleotide, two hydrogen bonds with the
backbone and side chain of Tyr865, and one hydrogen bond with Lys950 side chain.
4 DNA Dependent DNA Polymerases as Targets for Low-Weight …
• correlation analysis of the "structure-activity" and selecting the “compound hits”
for directed (purposeful) chemical optimization based on the results of biochemical assays;
• chemical synthesis and optimization of new inhibitors based on the analysis of
“structure–activity”, selectivity and computer simulation data.
From this viewpoint the clear understanding of structural aspects of protein–ligand
interaction is a key factor for the correct prediction of inhibitor affinity. Unfortunately, structural information about binding sites for non-nucleoside inhibitors on
the surface of DNA-dependent DNA polymerases is rather limited (see the previous
section).
At the same time, researchers start to use various computational approaches in
the rational design procedure of new (nucleoside analogs or nucleoside containing)
inhibitors of DNA-dependent DNA polymerases (especially, human pols). Spectrum of applicable approaches is sufficiently broad and includes all the above-mentioned methods of the structural analysis in silico.
Richartz and co-workers [157] have successfully applied the methods of molecular docking and molecular dynamics for the analysis of structural action mechanisms
of several nucleotide analogs—potential inhibitors of human DNA polymerase α
(pol α), and of the well-known non-nucleotide pol α inhibitor aphidicolin [56, 157].
In molecular dynamics simulations, aphidicolin occupied the catalytic centre,
but acted in a not truly competitive manner with respect to nucleotides. It destabilized the replicating “closed” form of the pol alpha and transferred the enzyme
into the inactive “open” conformation [157]. This result is consistent with recent
experiments on the binding mode of aphidicolin. Unfortunately, aphidicolin could
not be introduced into therapy because of its toxicity and rapid metabolism after
systemic application [158]. Among studied ‘nucleotides’, the highest potential for
selective pol α inhibition was established for 2-butylanilino-dATP (BuAdATP). The
butylphenyl moiety of BuAdATP occupies a lipophilic pocket, formed by the residues Leu960, Leu972, Val976, Ile869, Tyr865 and Tyr957. These lipophilic interactions, coupled with hydrogen bonds between BuAdATP, template nucleotide and
side chains of residues Tyr865 and Lys950, are likely to be responsible for the good
inhibition efficiency of the butylanilino derivatives. The lowest abilities to inhibit
human pol α were demonstrated for lamivudine-TP and zidovudine-TP (both compounds in three phosphate form) [157].
Later Höltje and co-workers (from the same research team) reported the development of several new human polymerase α inhibitors applicable for skin tumor
treatment (in order to design new drugs for actinic keratosis and squamous cell
carcinoma) [159] (Fig. 4.10). To study the binding modes of these compounds, the
same computational approaches as described in Richartz et al. [157] were used.
It was shown that the compound HM1-TP forms two hydrogen bonds with the
DNA template nucleotide, two hydrogen bonds with the side chain of Lys950 and
one hydrogen bond each with the backbone of Tyr865 and the side chain of Arg922
in the active center of human pol α. At the same time BuP-OH-TP forms two hydrogen bonds with the DNA template nucleotide, two hydrogen bonds with the
backbone and side chain of Tyr865, and one hydrogen bond with Lys950 side chain.
4 DNA Dependent DNA Polymerases as Targets for Low-Weight …