Interaction of tricyclic drug analogs with synaptic plasma membranes: structure-mechanism relationships in inhibition of neuronal Na+/K(+)-ATPase activity.

Perturbations of rat brain synaptic plasma membrane (SPM) bilayer structure and Na+/K(+)-ATPase activity were correlated for drugs that are structurally related and exhibit similar toxicological side effects but belong to different pharmacological classes. Na+/K(+)-ATPase IC50 values decrease linearly with increasing octanol/water partition coefficients (log-log plot) for a series of dimethylethylamine-containing drugs (i.e., chlorpromazine, amitriptyline, imipramine, doxepin, and diphenhydramine), emphasizing hydrophobicity in inhibition. However, nortriptyline and desipramine are 1.2 log units less hydrophobic than their N-methylated parent drugs but more potent inhibitors. To investigate this, bilayer surface structure was examined by the binding of the fluorophore 1-anilinonaphthalene-8-sulfonic acid (ANS) to SPMs. The dissociation constant and wavelength maximum of ANS are invariant with drug binding; however, the limiting fluorescence intensity of ANS (F infinity) is increased. Such data indicate that these cationic drugs bind to the membrane surface, increasing the number but not the polarity of ANS binding sites by cancelling charge at anionic phospholipid groups. More importantly, there is a close linear correlation between the concentrations of drugs necessary to increase F infinity by 40% and the IC50 values, with full compensation for the N-demethylated drugs. This correlation implies that drug-induced increases in SPM-bound ANS fluorescence are a better predictor of Na+/K(+)-ATPase inhibition than are octanol/water partition coefficients and that electrostatic interactions are also involved in inhibition. Furthermore, it points toward similar mechanisms of biomembrane surface interaction governing both inhibition and fluorescence change that are common to these drugs. K(+)-dependent p-nitrophenylphosphatase activity is inhibited with the same potency as Na+/K(+)-ATPase activity, indicating that inhibition may involve drug interaction near the K+ binding sites. Furthermore, chlorpromazine, diphenhydramine, and dimethylaminopropyl chloride alter K(+)-activation of K(+)-dependent p-nitrophenylphosphatase, progressing from noncompetitive through mixed to competitive inhibition as their hydrophobicity changes, and these mechanisms are consistent with steric hindrance of K+ binding. In contrast to the ANS data, decreases in 1,6-diphenyl-1,3,5-hexatriene fluorescence anisotropy induced by these drugs do not correlate with Na+/K(+)-ATPase inhibition, and drug N-demethylation enhances inhibition without altering anisotropy; both findings indicate that Na+/K(+)-ATPase activity is not predominantly influenced by changes in bulk fluidity. Taken together, these data suggest that electrostatic interactions at the biomembrane surface between the protonated amino group of the drug and anionic groups on the enzyme and/or phospholipids near the K+binding sites are crucial to inhibition and that drug hydrophobicity modulated the number and orientation of these interactions.