Phylogenetic value of aphidicolin-induced fragile sites and the evidence for the existence of anti-fragile sites in four species of deer mice

Aphidicolin (APC)-induced chromosomal breakage data were analyzed by alternative statistical models (FSM and FSM3) for per-individual identification of fragile sites within and among four species of Peromyscus. The FSM methodology classifies chromosomal loci into two categories, non-fragile and fragile. Alternatively, the FSM3 methodology assumes and accounts for the existence of a third statistical category of loci, anti-fragile sites.; To test for the assumption of anti-fragile sites, a data-based test was used to evaluate the hypothesis that non-fragile sites comprise only those loci which experience positive and random chromosomal breakage. A compilation of the chromosomal breakage pooled over the 40 individuals indicated no breakage at 55 (16.5%) of the 334 chromosomal bands. Assuming that breakage at non-fragile sites is statistically independent, the data were pooled and analyzed using Feller's Poisson Approximation. These results indicated a greater than 85% probability that the sample of 7759 breaks should have saturated the 334 bands. These results provide reasonable support for the conclusion that at least some of the 55 sites at which breaks were not observed are truly anti-fragile.; The FSM and FSM3 identifications of fragile sites within P. maniculatus, P. polionotus, P. keeni, and P. leucopus were consistent in indicating inter-individual variation in the distribution and number of fragile sites within each species. Among the four species of Peromyscus , the total number of different FSM-identified fragile sites ranged from 34–49. These data indicate that the number and specific composition of APC-induced fragile sites varies among individuals.; Data presented here is the first cladistic analysis of chromosomal fragile sites designed to examine the evolutionary relevance of these structures. Per-individual parsimony analyses of the data (FSM- and FSM-identified fragile sites) clearly falsify the hypothesis that APC-sensitive fragile sites are genetically irrelevant artifacts of clastogenic induction. Phylogenetic analyses of the per-species presence and absence of fragile sites yielded maximum parsimony trees that were entirely concordant with the corroborated phylogeny for the species examined. Evolution of APC-induced fragile sites is apparently quite rapid as the internal branches of both the FSM- and FSM3-based phylogenies were characterized by cladistically informative gains and losses of fragile sites.