The actions of two clinically important dibenzocycloheptane antidepressant drugs, amitriptyline and nortriptyline, were studied on ionic channels of nicotinic acetylcholine (AcCho) receptors at the neuromuscular junction of frog skeletal muscle. Amitriptyline (5-10 ,AM) and nortriptyline (1-2 !LM), like imipramine (5-10 jmM), did not react with the nicotinic AcCho receptor but caused a voltage-and time-dependent decrease in the peak amplitude of the endplate current (epc). The time constant ofepc decay, however, retained its voltage sensitivity. The voltageand time-dependent effect of amitriptyline was nonlinear with regard to the current/voltage (I/V) relationship. Nortriptyline also had a more pronounced voltage-and time-dependent effect evidenced by a hysteresis loop in the I/V relationship of the epc due to the drug's greater potency at more negative potentials. The nonlinearity and hysteresis loop in the I/V relationship of the epc was eliminated by the use of50-msec stepwise changes ofthe membrane potential. The nonlinearity and hysteresis were due to a time-dependent phenomenon and did not involve previous AcCho receptor activation. The rate constant of the voltage-and timedependent decrease in epc amplitude was sensitive to the membrane electric field and varied linearly with the membrane potential. lontophoretically elicited epcs were much more depressed by both drugs than were spontaneous miniature epcs. There was no effect on the time constant ofminiature epc decay, single-channel lifetime, or conductance. Thus (as we have pointed out in our histrionicotoxin studies) the primary site of action of these agents presumably is the activated but nonconducting species ofthe ionic channel of the nicotinic AcCho receptor. These agents, particularly nortriptyline, point to several different binding sites of the ionic channel and are suitable tools for the separation of the effects on peak current amplitude from its time constant of decay.(Perhydro)histrionicotoxin and phencyclidine produce a blockade of ionic channels associated with nicotinic acetylcholine (AcCho) receptors at concentrations that have no effect on the binding of either AcCho or a-bungarotoxin (1-3). Despite Recently, a study has revealed that four tricyclic antidepressant drugs-namely, imipramine, desipramine, amitriptyline, and nortriptyline-have the ability to bind to the ionic channel of the nicotinic receptor and thus inhibit the binding ofchannel probes, such as phencyclidine and perhydrohistrionicotoxin (9). This inhibition of channel probe binding was enhanced when channel activation was induced with AcCho, a process that probably involved only one site (9). Indeed, one of the tricyclic antidepressants, imipramine, decreased peak epc amplitude of frog sartorius muscle in a concentration-dependent manner (10, 11). Furthermore, with imipramine the I/V relationship of the epc became nonlinear, an indication of ionic channel blockade (12, 13), with negligible effect on Tepc. This decrease in peak epc amplitude by an ionic channe...