Multiple mechanisms of hERG liability: K+ current inhibition, disruption of protein trafficking, and apoptosis induced by amoxapine.

The antidepressant amoxapine has been linked to cases of QT prolongation, acute heart failure, and sudden death. Inhibition of cardiac hERG (Kv11.1) potassium channels causes prolonged repolarization and is implicated in apoptosis. Apoptosis in association with amoxapine has not yet been reported. This study was designed to investigate amoxapine effects on hERG currents, hERG protein trafficking, and hERG-associated apoptosis in order to elucidate molecular mechanisms underlying cardiac side effects of the drug. hERG channels were expressed in Xenopus laevis oocytes and HEK 293 cells, and potassium currents were recorded using patch clamp and two-electrode voltage clamp electrophysiology. Protein trafficking was evaluated in HEK 293 cells by Western blot analysis, and cell viability was assessed in HEK cells by immunocytochemistry and colorimetric MTT assay. Amoxapine caused acute hERG blockade in oocytes (IC(50) = 21.6 microM) and in HEK 293 cells (IC(50) = 5.1 microM). Mutation of residues Y652 and F656 attenuated hERG blockade, suggesting drug binding to a receptor inside the channel pore. Channels were mainly blocked in open and inactivated states, and voltage dependence was observed with reduced inhibition at positive potentials. Amoxapine block was reverse frequency-dependent and caused accelerated and leftward-shifted inactivation. Furthermore, amoxapine application resulted in chronic reduction of hERG trafficking into the cell surface membrane (IC(50) = 15.3 microM). Finally, the antidepressant drug triggered apoptosis in cells expressing hERG channels. We provide evidence for triple mechanisms of hERG liability associated with amoxapine: (1) direct hERG current inhibition, (2) disruption of hERG protein trafficking, and (3) induction of apoptosis. Further experiments are required to validate a specific pro-apoptotic effect mediated through blockade of hERG channels.

Quantitative prediction of catalepsy induced by amoxapine, cinnarizine and cyclophosphamide in mice.

Parkinsonism can be a side effect of antipsychotic drugs, and has recently been reported with peripherally acting drugs such as calcium channel blockers, antiarrhythmic agents and so on. In this study, we examined the quantitative prediction of drug-induced catalepsy by amoxapine, cinnarizine and cyclophosphamide, which have been reported to induce parkinsonism. Dose-dependent catalepsy was induced by these drugs in mice. In vivo dopamine D(1), D(2) and muscarinic acetylcholine (mACh) receptor occupancies by these drugs in the striatum were also examined. The in vitro binding affinities (K(i) values) of amoxapine and cinnarizine to dopamine D(1), D(2) and mACh receptors in rat striatal synaptic membrane were 200 and 2900 nM, 58.4 and 76.4 nM and 379 and 290 nM, respectively. Cyclophosphamide did not bind to these receptors at concentrations up to 100 microM. Twenty drugs, including those mentioned above, showed a significant correlation between the observed intensity of catalepsy and the values predicted with a pharmacodynamic model (Haraguchi K, Ito K, Kotaki H, Sawada Y, Iga T. Prediction of drug-induced catalepsy based on dopamine D(1), D(2), and muscarinic acetylcholine receptor occupancies. Drug Metab Disp 1997; 25: 675-684) based on in vivo occupancy of dopamine D(1), D(2) and mACh receptors. We conclude that occupancy of dopamine D(1) and D(2) receptors contributes to catalepsy induction by amoxapine and cinnarizine.

Aspirin acetylates the tricyclic antidepressant amoxapine spontaneously to N-acetylamoxapine in vitro and in vivo.

1. N-Acetylamoxapine is formed nonenzymically in vitro, and in mice, from amoxapine, a tricyclic antidepressant, and aspirin. 2. Formation of acetylamoxapine from amoxapine and aspirin in vitro was maximal at pH 5.0 since this pH optimized reactant solubilities as well as decreasing aspirin hydrolysis. 3. Formation of aceylamoxapine from amoxapine and aspirin in mouse stomachs was rapid, and the pH study indicates that the intestinal pH would favour formation even more. 4. Acetylamoxapine administered to mice produced the same CNS-related signs, leading to death, as with amoxapine, but much larger doses and longer time periods were required to elicit these effects. As brain and liver levels of amoxapine in animals dying from acetylamoxapine administration were less than half those found in animals given lethal doses of amoxapine, the toxicity in mice of acetylamoxapine may not be due solely to deacetylation of acetylamoxapine to the parent compound.

Biochemical evidence that high concentrations of the antidepressant amoxapine may cause inhibition of mitochondrial electron transport.

Overdosage with the antidepressant amoxapine causes metabolic acidosis and may lead to brain damage and death. To better understand the metabolic disturbances caused by amoxapine overdose, its effects on three simple systems were studied: growth of Saccharomyces cerevisiae, mitochondrial energy metabolism, and an electron transport system in microsomal membranes. Growth of yeast on all substrates except lactate was inhibited by amoxapine at 50-100 micrograms ml-1. Growth on lactate was observed at 200 micrograms ml-1 of amoxapine. In beef heart mitochondria, amoxapine at 100 micrograms ml-1 inhibited reactions involving large sections of the electron transport chain. Energy-linked reactions in submitochondrial particles were also inhibited. Electron microscopy showed some disruption of the mitochondrial internal structure by amoxapine and a change from orthodox to condensed conformation. Microsomal NADH-cytochrome b5 reductase was inhibited by amoxapine, but at higher amoxapine concentrations than mitochondrial reactions. The results suggest amoxapine disrupts reactions of membrane-associated enzyme complexes, and mitochondrial energy conservation may be one of the first systems affected. We speculate that lactic acid accumulation in patients with amoxapine overdose may be caused by loss of electron acceptor activity in tissues.

Chronic but not acute treatment with antidepressants enhances the electroconvulsive seizure response in rats.

The effects of the chronic administration of antidepressants on threshold electroconvulsive (ECS) seizures were evaluated in rats. Initially, tonic-clonic seizures were induced in 90% of the animals. Of those animals responding with tonic-clonic seizures, 42% had hindlimb extension (extensors); the remainder showed only hindlimb flexion (flexors). No alteration in the pattern of seizures was observed 24 hr after a single oral dose of any of the antidepressant drugs. The rats were then treated with a total of 20 consecutive daily doses of antidepressants and threshold electroconvulsive seizure responses were evaluated 24 hr after the last dose. A significantly greater percentage of rats responded with extensor seizures after chronic treatment with amoxapine, chlorimipramine, parglyine and mianserin. There was no change in the pattern of seizures of the rats treated chronically with desipramine, but the duration of the clonic phase was reduced. After a 7 day period free of drugs a significantly greater percentage of animals had extensor seizures in the groups treated with amoxapine, chlorimipramine, pargyline and desipramine but not mianserin. In the light of evidence that chronic treatment with antidepressants reduces the activity of norepinephrine- or isoproterenol-sensitive adenylate-cyclase, and that the norepinephrine system is an important endogenous anticonvulsant factor in electroconvulsive seizures, these results suggest that the same mechanism may mediate both the therapeutic and proconvulsant effects of the chronic administration of antidepressants.

Amoxapine in experimental psychopharmacology: a neuroleptic or an antidepressant?

2-Chloro-11-(piperazinyl)dibenz[b,f][1,4]-oxazepine (amoxapine) gives an unusual spectrum in psychopharmacological tests. Many of its effects are similar to those of neuroleptics: sedation, decrease in motor activity, catalepsy (which is, however, qualitatively different from that induced by classical neuroleptics), transitory suppression of avoidance reaction, antagonism of amphetamine induced toxicity in crowded mice and inhibition of stereotyped behavior induced by amphetamine in rats, and antagonism to various effects of apomorphine (stereotyped behaviour in rats, climbing behaviour, stereotyped behaviour and hypothermia in mice). At similar doses which produce the above mentioned effects, amoxapine also shows effects atypical for a neuroleptic, but which are relatively characteristic of antidepressants: antagonism of prochlorperazine-induced catalepsy in rats, inhibition of reserpine induced hypothermia in mice and enhancement of yohimbine toxicity in mice. The profile of this substance does not facilitate the anticipation of therapeutic effects in humans.

The neuropharmacological actions of amoxapine.

Amoxapine possesses a broad spectrum of psychotropic actions, including antidepressant and neuroleptic effects in animals. Antidepressant activity is characterized by its ability to inhibit tetrabenazine-induced depression, antagonize reserpine-induced hypothermia and enhance yohimbine lethality. Neuroleptic activity is demonstrated by the ability of amoxapine to decrease locomotor activity, induce ptosis and catalepsy, inhibit apomorphine gnawing and amphetamine stereotyped behavior and by characteristic changes in monkey discriminated avoidance behavior. The fact that punished responding in squirrel monkeys was present was present after repeated administration may indicate an anti-anxiety action of this drug. Evidence is offered that the conversion of the tertiary terminal nitrogen to a secondary amine may alter the pharmacologica properties of dibenzoxazepines in a similar way to the for the phenothiazines.