Effect of substrate organization on the activity and on the mechanism of gentamicin-induced inhibition of rat liver lysosomal phospholipase A1.

Aminoglycoside antibiotics, such as gentamicin, induce a lysosomal phospholipidosis in the kidney cortex of experimental animals and humans. In vitro, gentamicin binds to negatively charged phospholipids, such as phosphatidylinositol, and decreases the activity of lysosomal phospholipases towards a neutral phospholipid (phosphatidylcholine) included in lipid vesicles. The mechanism of such an inhibition was not unequivocally established. On one hand Mingeot-Leclercq et al. (Biochem Pharmacol 37: 591-599, 1988) observed that the activity of phospholipase A1 is modulated by the negative charges of the bilayer and that the inhibitory potency of gentamicin is inversely related to the phosphatidylinositol content of the vesicles, and therefore proposed that inhibition is due to charge neutralization. On the other hand, Hostetler and Jellison (J Pharmacol Exp Ther 254: 188-191, 1990) observed that the activity of phospholipase A1 is not modulated by the negative charges of the vesicles and that the inhibitory potency of gentamicin is directly related to the phosphatidylinositol content of the bilayer, and therefore proposed that inhibition is due to substrate depletion. However, the experimental designs of these two models differed in several respects such as the source (liver versus kidney) and nature of the enzyme (native lysosomal extract versus purified delipidated phospholipase A1), and the composition of lipid vesicles (those containing constant amounts of phosphatidylcholine and cholesterol, and inversely varying amounts of phosphatidylinositol and sphingomyelin versus those containing inversely related amounts of phosphatidylcholine and phosphatidylinositol only). In order to assess the nature of the differences between these models, we compared the activity of phospholipase A1 and its inhibition by gentamicin using only one source of enzyme, the rat liver lysosomal extract, and the two types of lipid vesicles as used in the above models. Our results showed that both models are true within the frame work of their respective experimental designs. However, since the composition of the lipid vesicles as well as the nature of the enzyme preparation (whole lysosomal extract) in the "charge neutralization" model is closer to in vivo conditions, we suggest that this model may be more relevant to the in vivo situation.