Quantitative inter-relationships between aflatoxin B1 carcinogen dose, indole-3-carbinol anti-carcinogen dose, target organ DNA adduction and final tumor response

Abstract

A number of recent studies have described inhibitor-mediated reductions in the covalent DNA binding and tumorigenicity of various carcinogens, in species such as rats, mice and rainbow trout (Salmo gairdneri). Since inhibitory effects have, in most cases, been reported after testing at one carcinogen and one inhibitor level only, the detailed relationships between carcinogen dose, inhibitor dose, in vivo DNA binding and final tumor response are not well understood in any species. To determine these relationships we have employed the trout model in a combined DNA binding/tumor dose-response protocol using ∼10 000 animals. Trout were pretreated with one of five different dose-levels of indole-3-carbinol (I3C), a naturally occurring anti-carcinogen found in cruciferous vegetables such as broccoli and cabbage. After 4 weeks, animals received the same dietary level of I3C for a further 2 weeks together with [³H]aflatoxin B1 (AFB1) in the dose-range 10–320 p.p.b. From tanks containing 150 animals (three tanks per I3C-AFB1 dose-point), 15 fish were selected at random in order to assess hepatic AFB1-DNA binding levels. Remaining animals were returned to control diet for determination of tumor response at 12 months. Linear increases in DNA binding occurred with dose of AFB1 at each I3C dose-level. Successive increases In I3C dose gave dose-related decreases in AFB1-DNA binding, resulting in a series of curves of decreasing slope. Shifts in DNA-binding slopes were compared quantitatively with horizontal displacements towards higher carcinogen dose in corresponding tumor dose-response curves. At I3C doses of ≤2000 p.p.m., the inhibitor-altered tumor response was predicted precisely by changes in dose received (DNA adducts formed) in the target organ. These data constitute the first direct evidence of pure anti-initiating activity by a natural anti-carcinogen found in human diet, where all animals were treated at the same time and under identical conditions of exposure in both DNA binding and tumor studies. The data are discussed further in view of (i) their implications for DNA binding carcinogenicity correlations and the concept of ‘molecular dosimetry’, and (ii) limitations in the current database on anti-carcinogenesis as regards in vivo potency information, particularly for ‘ambivalent modulators’ which exhibit both inhibitory and promotional activity.