AIDS patients are widely recognized to be much more susceptible than non- AIDS patients to the side effects of the arylamine antibiotics dapsone (DDS) and sulfamethoxazole (SMX), which are used to treat or protect against Pneumocystis carinii pneumonia (PCP). These arylamines produce a constellation of hematologic (aplastic anemia and methemoglobinemia) and hypersensitivity (fever, rash) reactions which, in AIDS patients, are often severe enough to require withdrawal of the drug, even though PCP is a frequently fatal opportunistic infection in AIDS. The hematologic toxicities can be directly linked to cytochrome P450-dependent oxidation of the arylamine to an arylhydroxylamine, which is further oxidized spontaneously, in a co-oxidation with hemoglobin (in erythrocytes), and possibly by myeloperoxidase (in neutrophils) to a nitroso intermediate, a highly reactive electrophile which covalently binds to proteins producing anemia, neutropenia, and hepatotoxicity, either directly or through an immune mechanism. Intracellular glutathione (GSH) normally affords protection against metabolically-generated electrophiles. However, over the past few years, it has been observed that HIV seropositive individuals are substantially deficient in GSH. Based on these considerations, we propose to: 1) determine in AIDS patients with PCP whether inhibition of cytochrome P450 with cimetidine will inhibit the oxidation of DDS and SMX (using pharmacokinetic techniques), and diminish the toxicity of the arylamines; 2) determine in AIDS patients with PCP whether supplementation of GSH with the precursor N-acetylcysteine will retard the second step in SMX oxidation to the putative toxic electyrophile and diminish toxicity. (This approach is inadvisable for DDS since methemoglobinemia will be promoted); and 3) identify the specific cytochrome P450 isozyme responsible for arylamine N-oxidation in human liver microsomes using specific antibodies and inhibitors, identify a clinically acceptable specific inhibitor in human liver microsomes, and evaluate its effects in vivo as in Specific Aim 1. These studies are linked to existing protocols in the ACTU in which DDS and SMX are administered.