Page 595 - The Toxicology of Fishes

Chemical Carcinogenesis in Fishes 575

exposure. Due to their well-established role as proven, complete carcinogens in laboratory exposures with
rodents (Roe and Waters, 1967) and fishes (Bailey et al., 1987b; Hawkins et al., 1995; Hendricks et al.,
1980) and their ubiquitous occurrence in sediments (Gardner et al, 1998; Malins et al., 1985; Myers et
al., 1994; Varanasi et al., 1989; Wirgin et al., 1994), with particularly enhanced levels in sediment extracts
from highly contaminated sites (Baumann et al., 1996; Malins et al., 1985; Vogelbein et al., 1990), the
PAHs have rightfully deserved to be the primary focus of attention in field carcinogenesis studies.
What is the evidence that fish are exposed to these environmental PAHs? This question has been addressed
by comparisons of spectra from sediment to those from stomach contents of benthic fishes. Results indicate
that PAHs of sediments are also found in analyses of stomach contents of benthic fishes (Myers et al.,
1987). Furthermore, a study that involved 27 different sites on the Pacific Coast from Alaska to southern
California demonstrated the widespread nature of the contaminant investigations. When exposed, fish
readily accumulate PAHs (Varanasi et al., 1987). Although the concentrations of parent compounds in
muscle and liver are low, this is now known to reflect prior metabolism by liver. Fluorescent aromatic
compounds (FACs) in bile are products of hepatic PAH metabolism, and this provides a means of estab-
lishing exposure and comparing uptake in fishes from various sites. FACs analysis has become a widely
used approach to demonstrate exposure and host response to PAHs (Myers et al., 1991; Wirgin et al., 1994).
As was covered earlier, the PAHs are procarcinogens and must be metabolized to the ultimate carci-
nogenic form. Analysis of xenobiotic metabolizing enzymes of English sole from contaminated and
reference sites showed differences in levels as a function of the site from which they were collected
(Collier and Varanasi, 1991). Furthermore, cohorts held in the lab and fed different diets revealed changes
in the enzyme parameters and in FACs in the bile. Varanasi et al. (1989) demonstrated the relevance of
the metabolism to carcinogenesis in the English sole. They exposed sole to the parent compound in the
laboratory and detected the ultimate carcinogenic form of BaP, the BaP-7,8-diol-9,10 epoxide, and syn-
as well as anti-BaP deoxyguanosine adducts. Subsequently, isolated hepatocyte preparations from sole
were used to investigate metabolism of tritium-labeled BaP (Nishimoto et al., 1992). The cells formed
conjugated metabolites and syn- and anti-BaP diol epoxide adducts similar to those formed by intact livers.
What is the evidence that metabolism in wild fish is leading to the formation of adducts? Because
adduct formation represents a key step in the initiation of the carcinogenic process, it became important
to determine whether fish collected from highly contaminated sites showed more adducts and whether
these fish would have greater numbers of adducts when compared to the same parameter in fish collected
from reference sites. The study by Varanasi et al. (1989) lent credence to the notion that, of the PAHs,
chrysene, BaP, and dibenz(a,h)anthracene were involved in the adduct formation. Interestingly, they also
conducted a comparison between Puget Sound and Boston Harbor in tumor-prone English sole and
winter flounder, respectively, and showed similar adduct formation. Stein et al. (1993) investigated the
formation and persistence of BaP and 7H-dibenzo(c,g)carbazole (DBC) adducts. The latter compound,
a nitrogen-containing aromatic carcinogenic to rodents but not studied in fish, was important, as it was
considered possibly the most significant of the PAHs in sediments. Both compounds formed adducts,
and these declined over the 28-day course of study with half-lives of 11 and 13 days, respectively. Further
32
refinements of the P post-labeling technique with respect to quantitative analysis of PAH adducts now
allows for molecular dosimetry. With carcinogens covalently bound to DNA, an accelerated approach
through toxicokinetics to get at the actual delivered dose at the molecular target (Stein et al., 1993) and
the endpoints detected are associated with increased exposure to environmental PAHs.
What is the evidence that the PAHs and host biochemical and molecular alterations are related to
morphologic alterations in liver, specifically hepatic neoplasms? The evidence comes from statistical
analyses. First, when 3 species were studied in the 27 sites on the Pacific coast, all showed significantly
higher lesion prevalence in the contaminated urban as opposed to the reference sites (Varanasi et al,
1987). Second, concentrations of PAHs, PCBs, DDT and derivatives, chlordanes, and dieldrin in sedi-
ment, stomach contents, and liver and FACs in bile were significant risk factors for the occurrence of
hepatic lesions, including lesions designated as neoplastic, preneoplastic, non-neoplastic but proliferative,
and degenerative/necrotic lesions, as well as hydropic vacuolation primarily of hepatic biliary epithelium
(Myers et al., 1994). Earlier work by this group (Myers et al., 1991) demonstrated significant and
consistent statistical associations between levels of aromatic hydrocarbons in sediment and hepatic
neoplasms using a logistic regression analysis. Another way to approach this question is to bring the