Ab initio modeling of competitive drug–drug interactions: 5‐fluorouracil dimers in the gas phase and in solution

Abstract

5-Fluorouracil (FU) is a halogen derivative of the uracil nucleoside widely used for treatment of solid tumors and in combination drug regimens for cancer chemotherapy. In drug delivery by biodegradable poly(ester-ether-ester)s, drug-drug interactions are competitive with respect to the drug-copolymer ones. The potential energy and free energy of association for a variety of FU-FU dimers (either H-bonded or stacked) were therefore studied in the gas phase to shed some light on their absolute and relative strength, because competition between these two kinds of interactions was recently proposed [McCarthy et al. Mol Phys 1997, 91, 513] and examined with a combination of methods [Hobza et al. J Phys Chem A 1998, 102, 6921] for uracil dimers. The effect of the computational level [Hartree-Fock/second order Møller-Plesset (HF/MP2) or B3LYP, including or not counterpoise corrections to the basis set superposition error] was examined as well as that of the use of polarization functions more diffuse than usual 6-31G * (0.25) more diffuse than usual. The MP2 level is necessary to describe stacking interactions, because B3LYP is inadequate for them, giving results very similar to the HF ones. The 6-31G * (0.25) basis set produces very favorable interaction energies, but it is prone to basis set superposition error (BSSE). The binding free energy is noticeably less favorable than the potential energy, due to the increase in vibrational entropy for the association process. There is a good linear relationship between S vib and the binding energy for the HF/6-31G * optimized structures. Estimates of the incidence of thermal corrections at the MP2 level were carried out on the MP2/3-21G and B3LYP/6-31G * structures, where possible. The solvation properties in chloroform or in water of the dimers kept rigid at their in vacuo geometries were examined in the polarizable continuum model framework.

Interplay of intra‐ and intermolecular H‐bonds for the addition of a water molecule to the neutral and N‐protonated forms of noradrenaline

“…Therefore optimizations and energy evaluations can be carried out at the same level, i.e., HF/6‐31G*, without resorting to extra single point calculations with diffuse functions. Moreover, optimizations are not recommended when making use of very diffuse functions, because the geometry could badly be affected 30, 31.…”

“…Since the NAdrH + geometries in the adducts (a) significantly differ from the geometry of the isolated compound (i),* the counterpoise correction was estimated also taking into account the geometry deformation, E def 11g, 31. With the usual definition, Δ E (CP)= E (AB)− E aa (A)− E aa (B), where the first superscript stands for the geometry (as in the adduct) and the second for the basis set (of the adduct), Δ E (CP) includes the energy related to the geometry deformation, because Δ E is defined with respect to the isolated partners: Δ E = E (AB)− E ii (A)− E ii (B).…”

5‐fluorouracil dimers in aqueous solution: molecular dynamics in water and continuum solvation*

“…The vibrational frequencies at the MP2/6‐31G* level for the dimers considered in Ref. 6, now available with our facilities, allowed computing of the zero point vibrational energy (ZPE) correction, scaled by the suitable factor, f 17, i.e., f ΔZPE, and the Gibbs free energy (at T =298 K, p =1 atm) in the rigid‐rotor harmonic‐oscillator approximation 18, Δ G corr =Δ E vac + f ΔZPE+ΔΔ H (0− T )− T ΔΔ S (0− T ), here reported and compared to the MP2/3‐21G 19 and B3LYP/6‐31G* 20 estimates of Ref. 6.…”

“…The elucidation of the solvation properties of FU dimers is interesting per se, due to the importance of this fluorinated pyrimidine antimetabolite, and because FU is a good representative of a system with multiple hydration sites, able furthermore to produce either H‐bonded or stacked dimers 6, as do DNA base pairs 7, 8. In a previous article 6, we determined the free energy in vacuo and in water (or chloroform) solution of a few FU dimers of various type, concluding, on the basis of continuum solvation calculations 9, that the solvent can play a remarkable role in stabilizing the FU dimers. However, the continuum model does not account for discrete hydrogen bonding interactions, which could well be established between the heteroatoms present in FU and water.…”

“…However, the continuum model does not account for discrete hydrogen bonding interactions, which could well be established between the heteroatoms present in FU and water. In addition, these discrete interactions can affect the dimer geometries themselves, which earlier 6 were conversely kept rigidly to their in vacuo values.…”