Page 553 - The Toxicology of Fishes

Chemical Carcinogenesis in Fishes 533

processes are qualitatively shared among fish and mammalian models, though important, oftentimes
species-specific differences have been observed as well. The rapid advancement of molecular biology is
accelerating the elucidation of these similarities and differences, and their underlying basis. Relatedly,
fish comprise by far the most diverse class of vertebrates and thus provide a vast resource for studies in
comparative biology (Powers, 1989); for example, studies of related species of fish that display marked
differences in susceptibility to chemically induced cancer may help identify critical phenomena underlying
carcinogen sensitivity in specific groups of organisms, including fish and humans. Conversely, the shared
evolutionary history of fish and humans has resulted in opportunities to study process fundamental to all
species (e.g., signal transduction) in organisms biochemically less complex than humans.
Herein, we provide a historical perspective on this subject and describe the state of the science of
chemical carcinogenesis in fishes, including discussions of specific classes of carcinogens, carcinogen
metabolism, and molecular and biochemical mechanisms of mutagenesis, tumorigenesis, and DNA repair.
We also describe current laboratory and field approaches for investigating carcinogenesis in fishes.
Portions of these discussions will be placed in a comparative context with mammalian models, for which
more information is oftentimes available. For detailed discussions of chemical carcinogenesis in mam-
mals, the reader is referred to Pitot and Dragan (1991).

Historical Perspective

As a result of the intensification and acceleration of laboratory and field investigations of neoplasia in
fish during the past three decades, the body of knowledge in this highly specialized facet of cancer
research has been much expanded in breadth and depth. The refined laboratory toolkit was deployed
improving epidemiology of neoplasia in fish and supporting the association between environmental
contamination and this form of chronic toxicity. In the laboratory, molecular biology coupled with refined
and directed biochemical studies have uncovered events and their modulation that enhance or diminish
tumor development. In the analysis of the pathogenesis of the disease, new methods to more critically
evaluate the role of specific lesions were deployed. In retrospect, one can recapitulate events as follows.

Early Discoveries

Prior to mid-20th century, neoplasms had been known to occur in fishes but information about them
was limited to a few anecdotal descriptive reports (Wellings, 1969). Among the first reports was a liver
cell tumor diagnosed as an adenoma in a blue shark (Prionace glauca) in 1908 (Schroeders). This
anecdotal era ended abruptly in the mid-1950s and early 1960s, when in Italy, France, and the United
States large epizootics of hepatic neoplasms began to be recognized at many hatcheries, almost exclu-
sively in rainbow trout (Oncorhynchus mykiss, then Salmo gairdneri). By 1965 it was established that
aflatoxins in pelleted hatchery foods were the cause of the liver tumors (Halver and Mitchell, 1967; see
review by Sinnhuber et al., 1977). These epizootics of liver neoplasms became the first instance in which
the occurrence of fish tumors led to the recognition of a new family of chemical carcinogens. The
epizootics also led to recognition that the rainbow trout is more sensitive to the carcinogenic effects of
aflatoxins than are other salmonids and most, if not all, mammals that have been tested to date.
Until 1963, salmonids were the only family of bony fishes in which liver tumors had been found, and
all significant epizootics of fish with liver tumors had come from hatcheries. Perhaps because it was
known that salmonids would develop tumors after exposure to appropriate compounds, this opened the
door for workers using the trout model to investigate factors that modulate carcinogenesis, a theme that
not only continues to be actively pursued but is also of fundamental importance in risk assessment.
As we will see, those investigating other species of fishes took different approaches, either sampling
many individuals from a particular field site or, as in the case of small aquarium fishes, using proven
mammalian carcinogens in laboratory exposures followed by serial bioassay to determine alterations.
The latter led to a more in-depth appraisal of the pathogenesis of neoplasms in liver and in other organs.
In 1963, Dawe et al. (1964) performed a survey of feral fishes from Deep Creek Lake, Maryland, and