The mechanism of activation of microsomal glutathione transferase in isolated liver cells by diisapropylidene acetone (phorone) was investigated. Phorone (1 mM) causes a time-dependent increase (up to 2.6-fold) in the glutathione transferase activity of microsomes isolated from treated hepatocytes. Since phorone reacts with sulfhydryl groups, the possibility that this compound activated microsomal glutathione transferase directly was studied. It was found that neither the activity of the purified enzyme nor that in isolated microsomes is affected by phorone. It has been suggested [Masukawa T and Iwata H, Biochem Pharmacol 35: 435-438, 1986] that activation of microsomal glutathione transferase by phorone in vivo is mediated through thiol-disulfide interchange involving oxidized glutathione (GSSG). It is shown here that the glutathione transferase activity of isolated microsomes, which was increased by the addition of 10 mM GSSG, can be decreased to the basal level with 0.1 M dithioerythritol. Dithioerythritol, on the other hand, only marginally decreases the glutathione transferase activity in microsomes isolated from phorone-treated hepatocytes. This finding argues against a role for thiol-disulfide interchange in the activation of the enzyme by phorone. Furthermore, the glutathione depletion caused by phorone does not seem to be responsible for activation per se, since other thiol depletors [e.g. diethylmaleate (DEM)] do not affect the activity of the enzyme. Immunoblot analysis of microsomes isolated from phorone-treated hepatocytes did not reveal any partial proteolysis which might have accounted for the activation. It is suggested that activation of microsomal glutathione transferase by phorone proceeds through a mechanism which might reflect an in vivo regulation of this enzyme. Additional compounds which have been shown to activate the microsomal glutathione transferase in vivo were also tested and significant activation was obtained with 1,2-dibromoethane (1.4-fold) but not with DEM or carbon tetrachloride. Activation was also obtained with 1-chloro-2,4-dinitrobenzene (CDNB) (1.6-fold) and to a small extent with t-butyl hydroperoxide (1.2-fold). The activation by 1,2-dibromoethane and CDNB is probably mediated through covalent binding, considering the known alkylating properties of these compounds. CDNB is the first substrate shown to activate the microsomal glutathione transferase implying that electrophilic compounds which are substrates can increase the rate of their own elimination by reacting with this enzyme. In addition, activation by t-butyl hydroperoxide indicates that oxidative stress can activate microsomal glutathione transferase.