Interactions of aminoglycoside antibiotics with negatively charged lipid layers: Biochemical and conformational studies

R Brasseur, G Laurent, JM Ruysschaert… - Biochemical …, 1984 - Elsevier
R Brasseur, G Laurent, JM Ruysschaert, P Tulkens
Biochemical pharmacology, 1984Elsevier
Previous studies [Laurent et al., Biochem. Pharmac. 31, 3861 (1982)] have demonstrated
that aminoglycoside antibiotics bind to negatively charged phospholipid bilayers and inhibit
the activity of lysosomal phospholipases. This inhibition also occurs in vivo in animal and
man. It is considered to be an early and significant step in the development of
aminoglycoside-induced nephrotoxicity. The binding of 6 aminoclycosides in current clinical
use (dibekacin, gentamicin, tobramycin, kanamycin A, amikacin and streptomycin) to …
Abstract
Previous studies [Laurent et al., Biochem. Pharmac. 31, 3861 (1982)] have demonstrated that aminoglycoside antibiotics bind to negatively charged phospholipid bilayers and inhibit the activity of lysosomal phospholipases. This inhibition also occurs in vivo in animal and man. It is considered to be an early and significant step in the development of aminoglycoside-induced nephrotoxicity. The binding of 6 aminoclycosides in current clinical use (dibekacin, gentamicin, tobramycin, kanamycin A, amikacin and streptomycin) to phosphatidylinositol has been studied by gel fiitration technique and by conformational analysis. Variation of the phosphatidylinositol content from 0 to 27% of total phospholipids causes a cooperative increase in aminoglycoside binding. At fixed phosphatidylinositol concentration, the binding of the different aminoglycosides is related to the number of aminogroups carried by the drug (viz., gentamicin > kanamycin A > streptomycin) and is largely, but not entirely dependent upon electrostatic interactions. Conformational analysis of the interaction of aminoglycosides with phosphatidylinositol monolayers was made by a step-wise computation approach. We first have taken into account the Vander Waals, torsional and electrostatic energies and we have calculated the hydrophobic and hydrophilic centers of each molecule. Assembly was then computed by successive association of one molecule of drug and up to 4 molecules of phosphatidylinositol. The calculated interaction energies varied from −8.5 kcal/mol (gentamicin) to −4.9 kcal/mol (amikacin) and −3.9 kcal/mol (streptomycin). Electrostatic interactions were observed between the phospho groups of phosphatidylinositol and up to 3 of the 5 aminogroups in gentamicin, dibekacin or tobramycin (N1, N2, N'6) but only 2 of the 4 aminogroups in kanamycin A or amikacin (N3, N'6). A higher energy of interaction was also associated with a deeper penetration of the drug into the monolayer, due to significant, albeit minor hydrophobic interactions. The insertion of gentamicin, dibekacin and tobramycin in the bilayer was deep enough to allow the inositol groups to freely move over the aminoglycoside. Since we previously showed a correlation between the binding of aminoglycosides and the phospholipase inhibition [Carlier et al., Antimicrob. Ag. Chemother. 23, 440 (1983)], this approach may contribute to the rational design of less toxic aminoglycosides.
Elsevier