Modeling of copper(II) complexes with the SIBFA polarizable molecular mechanics procedure. Application to a new class of HIV-1 protease inhibitors

We have previously developed a new family of organometallic complexes targeting the HIV-1 protease, an enzyme that is essential for viral maturation. Among these, two Cu2+ complexes C1 and C2 were synthesized from flexible ligands L1, N1-(4-methyl-2-pyridyl)-2,3,6-trimethoxybenzamide and L2, N2-(2-methoxybenzyl)-2-quinolinecarboxamide, respectively. These ligands, designed to fit the protease active site, were shown to form 2:1 complexes with Cu2+. To compare the relative stability of C1 and C2 and to study the energetic and structural aspects of such large Cu2+ complexes, we have extended the polarizable molecular mechanics procedure SIBFA to treat this cation. This was done by carrying out parallel ab initio (HF, MP2) as well as DFT computations. A first validation step was done on monoligated complexes, in which the SIBFA energy components were shown to correctly match their ab initio counterparts from an energy-decomposition procedure. Subsequent tests on polyligated Cu2+ complexes with neutral and anionic ligands showed the procedure to reproduce the results from MP2 and DFT computations with good accuracy (relative error <3%). We have next extended the calculations to compare the C1 and C2 complexes. Energy balances including continuum solvation effects indicate a greater stability of the C1 complex, in agreement with experimental results. The effects of the methoxyphenyl substituents on the C1 complex stability were further investigated and compared to their concomitant influences on the ab initio-computed molecular electrostatic potential.