G-quadruplex-binding and telomerase-inhibiting
capacities of G-quadruplex
ligands were examined under a cell nuclei-mimicking condition including
excess double-stranded DNA (λ DNA) and molecular crowding cosolute
(PEG 200). Under the cell nuclei-mimicking condition, a cationic porphyrin
(TMPyP4) did not bind to the G-quadruplex despite the high affinity
(<i>K</i><sub>a</sub> = 3.6 × 10<sup>6</sup> M<sup>–1</sup>) under a diluted condition without λ DNA and
PEG 200. Correspondingly, TMPyP4 inhibited telomerase activity under
the diluted condition (IC<sub>50</sub> = 1.6 μM) but not under
the cell nuclei-mimicking condition. In contrast, the <i>K</i><sub>a</sub> and IC<sub>50</sub> values of an anionic copper phthalocyanine
(Cu-APC) under the diluted (2.8 × 10<sup>4</sup> M<sup>–1</sup> and 0.86 μM) and the cell nuclei-mimicking (2.8 × 10<sup>4</sup> M<sup>–1</sup> and 2.1 μM) conditions were similar.
In accordance with these results, 10 μM TMPyP4 did not affect
the proliferation of HeLa cells, while Cu-APC efficiently inhibited
the proliferation (IC<sub>50</sub> = 1.4 μM). These results
show that the cell nuclei-mimicking condition is effective to predict
capacities of G-quadruplex ligands in the cell. In addition, the antiproliferative
effect of Cu-APC on normal cells was smaller than that on HeLa cells,
indicating that the cell nuclei-mimicking condition is also useful
to predict side effects of ligands.