Page 231 - The Toxicology of Fishes

Biotransformation in Fishes 211

(for a review, see Eaton and Gallagher, 1994). When the administered AFB dose is normalized to target
1
dose (e.g., AFBO–DNA adducts per 10 nucleotides), a highly linear relationship between DNA adduct
8
formation and tumor response is obtained, even when using combined data from both rats and rainbow
trout (Buss et al., 1990). Furthermore, a large range of linearity exists among total administered dose
and AFBO–DNA adduct levels in rainbow trout administered dietary AFB (Dashwood et al., 1988).
1
Collectively, these aforementioned studies demonstrating the linear relationship between AFB dose and
1
AFBO–DNA adducts are not supportive of a threshold hypothesis for aflatoxin genotoxicity at low doses,
at least not in two highly sensitive yet diverse species (rats and rainbow trout).
Studies of AFB metabolism and carcinogenesis have demonstrated that AFB biotransformation is
1
1
intimately linked with its toxic and carcinogenic effects. Accordingly, aquatic species differences among
AFB biotransformation pathways are a critical determinant underlying variations in species sensitivities
1
of AFB -induced carcinogenesis. Although rainbow trout are extremely sensitive to the hepatocarcinogenic
1
effects of AFB , Coho salmon (Oncorhynchus kisutch), a closely related salmonid, are resistant to AFB 1
1
carcinogenicity (Hendricks, 1994). The biochemical basis for this difference in AFB sensitivity among
1
salmonids may be due to less efficient CYP-mediated AFB epoxidation in the Coho salmon relative to
1
the trout. Specifically, AFB –DNA binding was reported to be 56-fold greater in rainbow trout liver than
1
in Coho salmon after intraperitoneal (i.p.) AFB administration and 18-fold greater after dietary AFB 1
1
exposure (Bailey et al., 1988). Other pathways, such as phase II metabolism and AFB elimination, are
1
relatively similar among the two species (Bailey et al., 1988), indicating that microsomal P450-mediated
AFB epoxidation and subsequent DNA binding accounts for differences in AFB –DNA binding among
1
1
the two salmonids. Like Coho salmon, microsomes prepared from channel catfish (Ictalurus punctatus)
liver are inefficient at catalyzing AFB epoxidation (Gallagher and Eaton, 1995). Channel catfish injected
1
with AFB show no elevation in DNA damage as detected by the comet assay, as opposed to rainbow
1
trout, which display extensive DNA damage in blood, liver, or kidney after exposure (Abd-Allah and el-
Fayoumi, 1999). Thus, the evidence to date suggests that the resistance of some fish to AFB carcinogenesis
1
can be attributed to inefficient conversion of the procarcinogen to the DNA-reactive metabolite.
The microsomal CYP-dependent monooxygenases also oxidize AFB to its hydroxylated metabolites,
1
AFM , AFP , and AFQ (Figure 4.23). AFQ is formed via 3α-hydroxylation of AFB , whereas AFM 1
1
1
1
1
1
is produced by 9α-hydroxylation of AFB . O-Demethylation of AFB results in the formation of AFP .
1
1
1
The acute toxicities of the hydroxylated metabolites are generally lower than the parent compound (Hsieh
et al., 1974; Stoloff et al., 1972), as are the mutagenic potencies (Coulombe et al., 1982, 1984; Hsieh
et al., 1984). However, AFP and AFQ do not appear to be major oxidative metabolites of AFB 1
1
1
metabolism in fish. In contrast, significant amounts of AFM are formed in fish (Ramsdell and Eaton,
1
1990; Ramsdell et al., 1991; Sinnhuber et al., 1974). Dietary AFM is approximately 30% as carcinogenic
1
as AFB in trout (Sinnhuber et al., 1974), whereas the carcinogenic potency of AFQ is approximately
1
1
1% that of AFB (Hendricks et al., 1980). Although other aflatoxin metabolites, including the epoxides
1
of AFM , AFP , and AFQ , may contribute to DNA binding, the evidence to date strongly indicates that
1
1
1
such secondary oxidation products are of minor importance (Bailey, 1994; Raney et al., 1992b).
CYP2K1 is the major salmonid P450 isozyme that activates AFB to AFBO (Williams and Buhler,
1
1983; Yang et al., 2000). Immunoquantitation studies of salmonid P450 isozymes indicate that Coho
salmon microsomes express less CYP2K1 than rainbow trout, thus providing a mechanistic basis for
the lack of AFB oxidation by Coho salmon (Bailey et al., 1988). Juvenile trout injected with the
1
estrogenic and androgen hormones 17β-estradiol and testosterone have lower mRNA and protein levels
of CYP2K1 and reduced AFB –DNA binding relative to control animals, suggesting that hormonal status
1
may affect the ability of rainbow trout to form the toxic AFBO (Buhler et al., 2000). A CYP2K1 ortholog
is also present in zebrafish (Danio rerio) (Troxel et al., 1997), and it is likely that this or a related isoform
is responsible for the activation of AFB in this species. Like channel catfish, rainbow trout treated with
1
BNF exhibit increased AFM production (Goeger et al., 1988). In general, the capacity for AFBO
1
stereoisomer production has not been measured in fish. It is reasonable to assume that species such as
zebrafish and rainbow trout, which can catalyze AFB –DNA binding, can form exo-AFBO as a significant
1
proportion of their microsomal AFBO. Furthermore, the fact that female zebrafish form more
AFBO–DNA adducts than do males may be a reflection of a higher ratio of AFB -exo/-endo epoxide
1
formation for females (Troxel et al., 1997). Interestingly, the level of AFBO–DNA adduct formation in