THE MODEL AH-RECEPTOR AGONIST BETA-NAPHTHOFLAVONE INHIBITS AFLATOXIN B-1-DNA BINDING IN-VIVO IN RAINBOW-TROUT AT DIETARY LEVELS THAT DO NOTINDUCE CYP1A ENZYMES

beta-Naphthoflavone (BNF), a well-known Ah-receptor agonist, has been believed to inhibit aflatoxin B-1 (AFB(1)) carcinogenesis in rats and rainbow trout primarily through induction of the cytochrome P450 1A (C YP1A) enzyme subfamily and consequent diversion of AFB(1) to the less carcinogenic phase I metabolite aflatoxin M(1) (AFM(1)). This study in vestigates the dose responsive effects of dietary BNF treatment on CYP 1A induction, AFM(1) formation, AFB(1)-8,9-epoxide formation and AFB(1 )-DNA binding in the trout model. Pre-feeding diet containing 10-200 p .p.m. BNF after AFB(1) i.p. injection provided dose-dependent inductio n of CYP1A-dependent ethoxyresorufin-O-deethylase (EROD) activity and inhibition of in vivo AFB(1)-DNA binding. However, most of the observa ble inhibition of DNA adduction (45% inhibition) had occurred at 10 p. p.m. BNF without detectable EROD induction; higher doses of BNF up to 200 p.p.m. induced EROD >6-fold but provided only another 15% inhibiti on of DNA adduction in vivo, When in vitro AFB(1)-DNA binding was asse ssed using liver microsomes from trout fed 10-100 p.p.m. BNF, induced microsomal EROD activity correlated moderately with reduction of in vi tro AFB(1)-DNA binding activity, However, BNF treatment in a low dose range (0.2-10 p.p.m.) also strongly inhibited in vivo hepatic AFB(1)-D NA binding (69% inhibition at 5 p.p.m. BNF in this experiment), in a d ose-dependent manner, in the complete absence of detectable EROD induc tion. The microsomes from 5 p.p.m. BNF-treated trout had no more EROD activity than control microsomes, and no less capacity for catalyzing AFB(1) - DNA binding in vitro than control microsomes. Thus, the poten t inhibition of hepatic AFB(1)-DNA binding ill vivo by 5 p.p.m. BNF wa s a result of neither CYP1A enzyme induction nor irreversibly reduced catalytic capacity for AFB(1)-8,9-epoxide formation, Direct analysis o f AFB(1) metabolites formed in vitro by liver microsomes from trout fe d 10, 100 and 500 p.p.m. BNF showed that low dietary BNF (10 p.p.m.) n either induced microsomal CYP1A-mediated AFM(1) formation nor altered AFB(1)-8,9-epoxide formation compared to the control. By comparison, 1 00 and 500 p.p.m. BNF pretreatment significantly elevated microsome-ca talyzed AFM(1) formation ill vitro (P < 0.001), and this increase was highly correlated with increased EROD activity (r(2) = 0.999, P < 0.00 1). Upon in vitro addition, BNF was found to be a potent inhibitor of microsome-mediated AFB(1)-8,9-exo-epoxide formation (IC50 = 2.6 +/- 0. 1 mu M) and AFB(1)-DNA binding (inhibition constant K-i = 3.03 +/- 0.2 5 mu M). These findings indicate that CYP1A enzyme induction can contr ibute modestly to BNF protection against AFB(1) in this species both i n vivo and in vitro at higher BNF doses, but does not do so at lower d oses. Instead, enzyme inhibition by BNF against AFB(1) 8,9-epoxidation appears to be the predominant protective mechanism at higher BNF dose s, and the sole protective mechanism at low doses, in the rainbow trou t, These findings demonstrate that mechanisms of chemoprevention can c hange with anticarcinogen dose, and caution that even potent induction of phase I or phase II activities does not assure that pathway to be a predominant protective mechanism in vivo.