Page 582 - The Toxicology of Fishes

562 The Toxicology of Fishes

expression of Tu and development of malignant melanomas. Although phenotypic expression of Tu (as
measured by the number and size of spots and the malignancy of the melanomas) highly correlated with
elevated expression of the c-src gene (Schartl et al., 1985), further studies revealed that Tu and src were
different genes (Maueler et al., 1988).
Restriction fragment length polymorphism (RFLP) analysis of a Xiphophorus v-erbB homolog indi-
cated that this oncogene cosegregated with sex chromosome-linked melanoma in Xiphophorus hybrids
(Adam et al., 1988; Zechel et al., 1988). Partial sequence analysis revealed not one, but two, distinct
Xiphophorus genes (designated by Anders and colleagues as x-egfrA and x-egfrB) with homology to the
human (c-erbB) gene. The most highly conserved region was the tyrosine kinase domain. A nearly full-
length cDNA clone was isolated and sequenced from a melanoma-derived cell line and, as with the
x-egfrB and x-egfrA genes above, the clone (designated Xmrk, for Xiphophorus melanoma receptor kinase
gene) showed significant homology to the human EGF receptor (Wittbrodt et al., 1989). The Xmrk protein
also induces downstream signaling cascades, including ras/raf and MAP kinase, in a manner similar to
the mammalian homolog (Meierjohann et al., 2004). Another downstream interaction identified is with
focal adhesion kinase (FAK), which, in turn, triggers pigment cell migration and partly explains the fast
metastasis observed in malignant melanoma tumor development (Meierjohann et al., 2006).
The difference in nomenclature adds to the confusion inherent in this system. The Xiphophorus
melanoma-associated gene related to the human c-erbB gene has been designated the Tu complex,
x-egfrB, x-erbB*, and Xmrk (Zechel et al., 1992). Nonetheless, the data are consistent with the inter-
pretation that Xmrk is the Xiphophorus tumor gene Tu, a novel receptor tyrosine kinase. Genetic analysis
indicated that the oncogenic potential of Xmrk was regulated by the Diff gene, which is postulated to
act as a tumor suppressor gene. Transgenic studies involving microinjection of Xmrk into medaka
embryos resulted in embryonic tumors (Dimitrijevic et al., 1988). In addition to overexpression, recent
studies have reported that two amino acid changes in the extracellular domain of the protein are also
sufficient to convert Xmrk into an oncogenic version leading to a constitutive, ligand-independent protein
(Winnemoeller et al., 2005). Diff has been determined to be a member of the CDKN2 family, as revealed
by studies of UVB-inducible melanoma in Xiphophorus hybrids (Nairn et al., 1996). This is significant
in that CDKN2A (mammalian homolog p16) is a tumor suppressor gene involved in human melanoma.
The Xiphophorus model remains perhaps the oldest and best-studied model of oncogene/suppressor
gene interaction. Of the model’s many unique aspects, the interaction of oncogenes (e.g., x-erbB*) and
suppressor genes (e.g., Diff) is of prime importance. This is currently the only vertebrate system where
intimate knowledge of oncogene and suppressor gene interactions and their molecular genetics appears
readily accessible. To date, a great many oncogenes and now tumor suppressor genes have been described
in Xiphophorus compared to other fishes. In addition to those discussed above, reports of abl, src, yes,
fyn, erbA, fes, fgr, fms, fos, int, ras, myb, sis, mil, myc, hck, lck, and yes have also appeared in the peer-
reviewed literature, and a number of other complete and partial sequences have been deposited with
GenBank (Adam et al., 1988; Maueler et al., 1988; Zechel et al., 1988).

Modifying Factors in Chemical Carcinogenesis
Earlier, we described the multiple stages of carcinogenesis, focusing on liver. Factors may modify
(enhance or inhibit) carcinogenesis and eventual liver tumor formation by affecting the uptake, transport,
and metabolism of the potential carcinogen, leading to changes in DNA adduct formation, or they may
affect the promotion or progression of the tumor. When AFB , aflatoxicol (AFL), aflatoxin M (AFM ),
1
1
1
and aflatoxicol M (AFLM ) were exposed to trout and the resultant molecular dosimetry (DNA adduc-
1
1
tion) was established, all aflatoxin adducts except those from AFLM were equally tumorigenic (Bailey
1
et al., 1988). In this instance, differences in tumor incidence with the various compounds were largely
or entirely accounted for by differences in uptake and metabolism leading to DNA adduction, rather
than inherent differences in tumor initiating potency per DNA adduct.
The rainbow trout model has been the principal source of our information regarding modifying factors
in fish exposed to chemical carcinogens. Production of hepatic tumors by single chemical agents has
been studied in detail, and a great body of evidence has been generated regarding tumor formation and