We have cloned, sequenced, and characterized the RNA expression properties of a ®sh CDKN2 gene from Xiphophorus helleri and X. maculatus. This gene, termed CDKN2X, shows a high degree of amino acid sequence similarity to members of the mammalian CDKN2 gene family, which includes the tumor suppressor loci CDKN2A (P16) and CDKN2B (P15). Comparative sequence analysis suggests that ®sh CDKN2X is similarly related to all four mammalian gene family members, and may represent a descendant of an ancestral prototypic CDKN2 gene. CDKN2X was mapped to a region on autosomal Xiphophorus linkage group V (LG V) known to contain the DIFF gene that acts as a tumor suppressor of melanoma formation in X. helleri/X. maculatus backcross hybrids. Thus, CDKN2X may be a candidate for the tumor suppressor DIFF gene. Here we have sequenced CDKN2X in both Xiphophorus species and have characterized its expression in normal and melanotic tissues within control and backcross hybrid ®sh. A simultaneous expressional analysis of the Xmrk-2 tyrosine kinase receptor gene, which is strongly implicated in melanomagenesis in this system, was also performed. RT ± PCR analyses revealed that both genes were highly expressed in melanomas. For CDKN2X, this result contrasts numerous ®ndings in human tumors including human melanoma in which either CDKN2A (P16) deactivation or LOH was observed.

Xiphophorus interspecies hybrids provide genetically controlled models of tumor formation. Spontaneous melanomas form in first-generation backcross (BC(1)) hybrids produced from backcrossing F(1) hybrids derived from the platyfish X. maculatus Jp 163 A and the swordtail X. helleri to the X. helleri parental strain (the Gordon-Kosswig hybrid cross). Nodular melanomas originate in the dorsal fin from cells constituting the spotted dorsal (Sd) pigment pattern. A parallel genetic cross, with X. maculatus Jp 163 B, exhibits the spotted side (Sp) pigment pattern instead of Sd, and produces BC(1) hybrids exhibiting a much lower frequency of spontaneous melanoma formation. These hybrids are susceptible to melanoma development if irradiated with UV light as fry. Other hybrids involving these two strains of X. maculatus and different swordtail and platyfish backcross parents also have been investigated as potential tumor models, and show differing susceptibilities to UV-induced and spontaneous melanomas. Genotyping of individual BC(1) hybrids from several Xiphophorus crosses has implicated a locus, CDKN2X (a Xiphophorus homologue of the mammalian CDKN2 gene family, residing on Xiphophorus linkage group V), in enhancing pigmentation and the susceptibility to spontaneous and UV-induced melanoma formation in BC(1) hybrids from some crosses, but not others. Homozygosity for X. helleri and X. couchianus CDKN2X alleles in BC(1) hybrids can predispose individuals to melanoma, but this susceptibility is modified in other crosses depending both on the contributing sex-linked pigment pattern locus from X. maculatus (Sd or Sp), and the genetic constitution of the backcross parent. Xiphophorus BC(1) hybrids constitute unique genetic models offering the potential to analyze the contributions of specific genes to spontaneous and induced tumor formation in different, but comparable genetic backgrounds.

The explosive expansion of gene maps of mouse and man has provided strong support for hypotheses first advanced from comparing fish and mammalian genomes that the vertebrate genome was derived from multiple ancestral tetraploidizations with subsequent preferential translocations among paralogous chromosomes. At least two genome duplication events have become widely accepted in lineages leading to vertebrates, and a third has been proposed either before, or after, divergence of fishes and tetrapods. Cytogenetic and comparative gene mapping studies suggest that teleost gene maps have diverged more slowly from gene arrangements in the vertebrate ancestor than have those of mammals. The recent assembly of extensive maps of >100 genes in three fish species, medaka (Beloniformes), Xiphophorus swordtails and platyfishes (Cyprinodontiformes), and zebrafish (Cypriniformes) and the development of less extensive maps in several other fish orders provides the first salient opportunity to assess homology of most or all chromosomes among fishes.