Pharmacophoric sites of anticancer metal complexes located using quantum topological atomic descriptors

Abstract In this paper, we present a new set of topological atomic descriptors (TAD) to identify pharmacophoric sites, using the anti-tumoral activities of metal copper (II) complexes. To this end, multiple linear regression-based models were built using TAD. These descriptors are related with the atomic electronic population, localization and delocalization, dipolar moment and quadrupole moment, and describe the way in which the atoms donate, accept or share electron density and how the electron density is polarized within an atomic domain. The regression models were built to reproduce the antiproliferative activity on two tumor human cells lines. Usually, the activity analysis of a metal complex focusses on the metal center features, but the ligands have a definitive role on the molecular recognition of the complexes by the biomolecules. In this case, the best models selected are those involving the changes of the delocalization and dipole moments of atoms within the C–N bonds of the ligands. These changes correlate with the antiproliferative activity and, thus, it allows to identify the C–N bonds as the pharmacophoric site of the copper complexes. These bonds are determinant for the recognition site of copper complexes by DNA backbone, as found in a previous molecular dynamics study performed by our group.