N-omega-Allyl-L-arginine is shown to be a competitive reversible inhibitor and time-dependent inactivator of bovine brain nitric oxide synthase (nNOS). The enzyme is protected against inactivation by the presence of the substrate, the absence of O-2, Or if NADP(+) is substituted for NADPH. The NADPH absorption spectrum is converted to that of NADP(+) concomitant with inactivation. The latter two results indicate that redox chemistry is required for inactivation. N-omega-Allyl-N-omega-hydroxy-L-arginine is synthesized and shown also to be a competitive inhibitor and time-dependent inactivator of nNOS, suggesting that it is a viable intermediate in the inactivation process. Inactivation of nNOS with N-omega-allyl-[C-14]-L-arginine or N-omega-[H-3]allyl-L-arginine followed by gel filtration or dialysis results in no radioactivity bound to the enzyme. It is shown spectrophotometrically as well as by HPLC that the heme is modified to four different species during inactivation. Inductively coupled plasma atomic emission spectrophotometry is used to show that 1 equiv of ferric ion is present in the modified hemes. When the heme is isolated after inactivation by the two radiolabeled inactivators, it is found that no C-14 is associated, but 0.9 equiv of H-3 is bound to the heme. This indicates that only the allyl part of the inactivator is bound to the heme. HPLC-electrospray mass spectrometry is used to show that the four modified hemes have the same mass which corresponds to heme plus an allyl group plus a hydrogen. The fact that the modified hemes no longer have an absorption at 400 nm but, instead, absorb at 280 nm suggests that four reduced and allylated hemes are produced (such as 34 or 35). N-omega-Propyl-L-arginine also is shown to be a competitive inhibitor and time-dependent inactivator of nNOS; inactivation requires O-2 and NADPH, and the substrate protects the enzyme from inactivation. Twenty-six equivalents of citrulline and nitric oxide is produced during inactivation with N-omega-propyl-L-arginine. This suggests that the double bond of the allyl group is not important to the inactivation mechanism. Possible mechanisms that rationalize these results are suggested. No N-omega-allyl-L-citrulline (39) is detected from the inactivation of nNOS by N-omega-allyl-L-arginine, which indicates that initial hydroxylation of the guanidino imine nitrogen does not occur. The only radioactive metabolite generated from N-omega-allyl-[C-14]-L-arginine inactivation is citrulline (13 equiv). Approximately 13 equiv of nitric oxide also are generated. When N-omega-allyl-N-omega-hydroxy-L-arginine is the inactivator, about 20 equiv of citrulline and nitric oxide is produced. When N-omega-[H-3]allyl-L-arginine is the inactivator, approximately equal amounts of (H2O)-H-3 and acrolein are produced (8-9 equiv); arginine also is a product. Acrolein does not inactivate nNOS. These last few findings support cleavage of the alpha-C-H bond of the allyl group as a turnover pathway which does not lead to inactivation.