Tacrine's [1,2,3,4-tetrahydro-9-acridinamine monohydrochloride monohydrate, (THA)] metabolic fate was examined using human and rat liver microsomal preparations. Following 1-hr incubations with human microsomes, [14C]THA (0.4 microM) was extensively metabolized to 1-hydroxyTHA with trace amounts of 2-, 4-, and 7-hydroxyTHA also produced. Poor recovery of radioactivity in the postreaction incubates suggested association of THA-derived radioactivity with precipitated microsomal protein. After exhaustive extraction, 0.034, 0.145, 0.126, and 0.012 nmol eq bound/mg protein/60 min of THA-derived radioactivity was bound to human liver preparations H109, H111, H116, and H118, respectively. Preparations H109 and H118 were lower in P4501A2 content and catalytic activity as compared with preparations H111 and H116. Incubations of equimolar [14C]1-hydroxyTHA with human liver microsomes also resulted in binding to protein, although to a lesser extent than observed with THA. [14C]THA (0.4 microM) was incubated for 1 hr with rat liver microsomes (1 microM P-450) prepared from noninduced (N), phenobarbital (PB), isoniazid (I), and 3-methylcholanthrene (3-MC)-pretreated animals. In all incubations, 1-hydroxyTHA was the major biotransformation product detected. After exhaustive extraction, 0.048, 0.054, 0.049, and 0.153 nmol eq/mg protein/60 min of THA-derived radioactivity was bound to microsomal protein from N, PB, I, and 3-MC pretreated rats. Increased binding with 3-MC induced rat liver preparations suggests the involvement of the P-450 1A subfamily in THA bioactivation. Glutathione (5 mM) coincubation inhibited the irreversible binding of THA-derived radioactivity in both human and 3-MC-induced rat liver preparations, whereas human epoxide hydrase (100 micrograms/incubate) had a relative minor effect. A mechanism is proposed involving a putative quinone methide(s) intermediate in the bioactivation and irreversible binding of THA. A species difference in THA-derived irreversible binding exists between human and noninduced rat liver microsomes, suggesting that the rat is a poor model for studying the underlying mechanism(s) of THA-induced elevations in liver marker enzymes found in clinical investigations.